Resource Manual on Flash Flood Risk Management Module 3: Structural Measures

Transcription

Resource Manual on
Flash Flood Risk
Management
Module 3: Structural Measures
1
About ICIMOD
The International Centre for Integrated Mountain Development, ICIMOD, is a regional knowledge
development and learning centre serving the eight regional member countries of the Hindu Kush
Himalayas – Afghanistan, Bangladesh, Bhutan, China, India, Myanmar, Nepal, and Pakistan – and
based in Kathmandu, Nepal. Globalization and climate change have an increasing influence on
the stability of fragile mountain ecosystems and the livelihoods of mountain people. ICIMOD aims
to assist mountain people to understand these changes, adapt to them, and make the most of new
opportunities, while addressing upstream-downstream issues. We support regional transboundary
programmes through partnership with regional partner institutions, facilitate the exchange of
experience, and serve as a regional knowledge hub. We strengthen networking among regional
and global centres of excellence. Overall, we are working to develop an economically and
environmentally sound mountain ecosystem to improve the living standards of mountain populations
and to sustain vital ecosystem services for the billions of people living downstream – now, and for
the future.
ICIMOD gratefully acknowledges the support of its core and programme donors:
the Governments of Afghanistan, Austria, Bangladesh, Bhutan, China, Germany, India, Myanmar,
Nepal, Norway, Pakistan, Sweden, and Switzerland, and the International Fund for Agricultural
Development (IFAD).
2
Resource Manual on
Flash Flood Risk
Management
Module 3: Structural Measures
Arun Bhakta Shrestha
Ezee GC
Rajendra Prasad Adhikary
Sundar Kumar Rai
International Centre for Integrated Mountain Development, Kathmandu, 2012
i
Published by
International Centre for Integrated Mountain Development
GPO Box 3226, Kathmandu, Nepal
Copyright © 2012
International Centre for Integrated Mountain Development (ICIMOD)
All rights reserved. Published 2012
ISBN 978 92 9115 265 0 (printed)
978 92 9115 266 7 (electronic)
LCCN2012-323209
Cover photo: Sundar Kumar Rai
Printed and bound in Nepal by
Hill Side Press (P) Ltd., Kathmandu, Nepal
Production team
Andrea Perlis (Senior editor)
A Beatrice Murray (Consultant editor)
Amy Sellmyer (Proofreader)
Dharma R Maharjan (Layout and design)
Asha Kaji Thaku (Editorial assistant)
Note
This publication may be reproduced in whole or in part and in any form for educational or non-profit purposes
without special permission from the copyright holder, provided acknowledgement of the source is made. ICIMOD
would appreciate receiving a copy of any publication that uses this publication as a source. No use of this
publication may be made for resale or for any other commercial purpose whatsoever without prior permission in
writing from ICIMOD.
The views and interpretations in this publication are those of the author(s). They are not attributable to ICIMOD and
do not imply the expression of any opinion concerning the legal status of any country, territory, city or area of its
authorities, or concerning the delimitation of its frontiers or boundaries, or the endorsement of any product.
This publication is available in electronic form at www.icimod.org/publications
Citation: Shrestha, AB; GC, E; Adhikary, RP; Rai, SK (2012) Resource manual on flash flood risk management –
Module 3: Structural measures. Kathmandu: ICIMOD
ii
Contents
Foreword
v
About This Module
vi
Acknowledgementsvii
Acronyms and Abbreviations
viii
Some Key Terms
ix
Chapter 1: Introduction
1
Major Factors Contributing to Flash Floods
1
Flood Risk Management 3
The Resource Manual
4
Chapter 2: Landslides and Debris Flows
5
What is a Landslide?
5
Mechanism of Landslide Formation
7
Factors Contributing to the Development of Landslides
8
Mitigation Measures 8
Chapter 3: Landslide Dam Lakes and Glacial Lakes
11
Landslide Dam Lakes 11
Glacial Lakes
12
Chapter 4: Bioengineering Measures
15
Functions of Vegetation
15
Bioengineering in Flash Flood Risk Management
16
Common Bioengineering Techniques in the Hindu Kush Himalayas
19
Chapter 5: Physical Methods for Slope Stabilization and Erosion Control 30
Terracing30
Diversions36
Grassed Waterways 38
Conservation Ponds
38
Retaining Walls 44
Drop Structures
49
Sabo Dams
54
Gully Control 54
Trail Improvement 58
iii
Chapter 6: Physical Methods for River Training Transversal Protection Structures
60
Longitudinal Protection Structures
67
Other Protection Structures
74
Chapter 7: Structural Measures for Flood Management in the
Context of Integrated Water Resources Management 78
Integrated Flood Management
78
Passive Flood Control Measures 79
Flood Storage Reservoirs
79
River Corridor Rehabilitation and Restoration 81
References iv
60
83
Foreword
The Hindu Kush Himalayan (HKH) region is endowed with plentiful natural resources, including an abundant
amount of water. At the same time, the region is prone to a wide range of water-related and other natural hazards,
including landslides, avalanches, flooding, and earthquakes, which result from the unstable geological conditions
and steep topography combined with often extreme weather. Among these, flash floods are particularly challenging
for communities.
Flash floods are severe flood events that occur with little warning. They can be triggered by intense rainfall,
extremely rapid snow melt, failure of natural or artificial dams, and outbursts of glacial lakes. The frequent
occurrence of flash floods in the Hindu Kush Himalayan region poses a severe threat to lives, livelihoods, and
infrastructure, both in the mountains and downstream. Vulnerable groups such as the poor, women, children, the
elderly, and people with disabilities are often the hardest hit. Flash floods tend to carry much higher amounts of
debris with them than normal floods and as a result cause more damage to hydropower stations, roads, bridges,
buildings, and other infrastructure.
Since its establishment in 1983, ICIMOD has explored different ways to reduce the risk of disasters from natural
hazards and the physical and social vulnerability of the people in the region. Approaches have included training
courses, hazard mapping, and vulnerability assessments as well as fostering dialogue among stakeholders and
developing materials for capacity building.
ICIMOD, in collaboration with various partners, has compiled and published resource materials on flash flood
risk management in order to support capacity development, and especially to support the training of planners and
practitioners. In 2008, ICIMOD published two modules of a resource manual on flash flood risk management
which focused on community-based management and non-structural measures. These were followed by a training
of trainers manual based on the resource materials. The present publication is the third module of the resource
manual and deals with structural measures. The publication was produced as a part of the project ‘Flash Flood
Risk Reduction – Strengthening Capacity in the Hindu Kush Himalayas’, supported by the United States Agency for
International Development, Office for Foreign Disaster Assistance (USAID/OFDA). We hope that this new volume
will be useful for practitioners and will make a meaningful contribution towards reducing disaster risk and providing
greater physical security for the people of this vulnerable region.
David Molden
Director General, ICIMOD
v
About This Module
Flash floods are among the most destructive natural disasters in the Hindu Kush Himalayan region. They are sudden
events that allow very little time to react, and often occur in remote and isolated mountain catchments where few,
if any, institutions exist that are equipped to deal with disaster mitigation, and where relief agencies are either
absent or have limited presence and capacity. Flash flood mitigation is generally addressed by community-based
organizations, local non-governmental organizations, or district and local level staff in government organizations.
But these groups often lack adequate understanding of the processes causing flash floods and knowledge of flash
flood risk management measures. Building the capacity of people working directly in flash-flood-prone catchments
will help to reduce the flash flood risk in the region.
The Resource Manual on Flash Flood Risk Management provides the materials needed to help people working in
risk prone areas understand the problem and manage the risk. The manual is published in three parts. The first
and second modules (published in 2008) focused on community-based management and non-structural measures
for managing flash flood risk. The present, third, module focuses on structural measures for flash flood risk
management. The material is divided into seven chapters: Chapter 1 describes the types of flash flood that occur
in the region and the most important contributing factors and summarizes the major approaches in flash flood risk
management; Chapters 2 and 3 describe measures to manage the risk from landslides and debris flows and from
landslide dam lakes and glacial lakes; Chapters 4 to 6 discuss bioengineering techniques, physical measures for
slope stabilization and erosion control, and physical measures for river training; and finally Chapter 7 presents the
concept of integrated flood management as a component of integrated water resource management. The manual
emphasizes throughout that structural measures are most effective and sustainable when implemented together with
appropriate non-structural measures.
The resource manual is aimed at junior to mid-level professionals with a civil engineering background. The aim is to
build the capacity of people working in district-level government and non-governmental organizations on flash flood
risk management. The measures described are simple yet effective. They were selected on the basis of being able to
be implemented using local and low-cost materials with a minimum of external materials and technical support, and
having a low environmental impact.
vi
Acknowledgements
This manual is an output of the project ‘Flash Flood Risk Reduction – Strengthening Capacity in the Hindu Kush
Himalayas’, which was supported by the United States Agency for International Development, Office for Foreign
Disaster Assistance (USAID/OFDA).
Our sincere thanks go to Jeebachh Yadav and Gita Khatri for their valuable contributions to the manual; to
Professor Hubert Holzmann of the University of Natural Resources and Applied Life Sciences, Austria, and Professor
Marco Borga of the University of Padua, Italy, for their very useful suggestions; to the Department of Water
Induced Disaster Prevention (DWIDP), Nepal, for providing essential resources; and to all those who contributed
photographs.
We are grateful to the many ICIMOD colleagues who contributed to the manual, especially Rajesh Thapa and
Sagar Ratna Bajracharya who provided important support and helped in preparing the draft document. Our thanks
go especially to Professor Hua Ouyang, then Programme Manager for Integrated Water and Hazard Management,
for seeing through the completion of the module, and to all the staff in this programme who supported its
preparation.
Finally, we would like to express our heartfelt thanks to the many other colleagues, both within and outside ICIMOD,
who read the manuscript and provided valuable comments and suggestions.
vii
Acronyms and Abbreviations
viii
BCM
Banked cubic metre
c/c
Centre-to-centre
GLOF
Glacial lake outburst flood
HKH
Hindu Kush Himalayas
IFM
Integrated flood management
ICIMOD
IWRM
Integrated water resource management
LCM
Loose cubic metre
LDOF
Landslide dam outburst flood
Some Key Terms
Bioengineering
The application of engineering design and technology to living systems; in terms of
natural hazard mitigation, it refers to the combination of biological, mechanical, and
ecological concepts to reduce or control erosion, protect soil, and stabilize slopes using
vegetation or a combination of vegetation and construction materials
Debris flow
A rapid mass movement of loose soil, rocks, and organic material along with entrapped
air and water forming a slurry that flows downslope
Disaster A serious disruption of the functioning of a community or a society causing widespread
human, material, economic, or environmental losses which exceed the ability of the
affected community or society to cope using its own resources
Disaster mitigation
Steps taken to contain or reduce the effects of an anticipated or already occurred
disaster; involves various levels such as individuals, groups, and communities; actions
taken depend in part on the perceptions of risk of those exposed
Disaster preparedness The knowledge and capacities developed by governments, professional response and
recovery organizations, communities, and individuals to effectively anticipate, respond to,
and recover from the impacts of likely, imminent, or current hazard events or conditions
Disaster risk
The potential disaster losses, in lives, health status, livelihoods, assets, and services,
which could occur to a particular community or a society over some specified future time
period
Disaster risk reduction The concept and practice of reducing disaster risks through systematic efforts to analyse
and manage the causal factors, including through reduced exposure to hazards,
lessened vulnerability of people and property, wise management of land and the
environment, and improved preparedness for adverse events
Factor of safety For a rock slope, the ratio of the total force resisting sliding to the total force tending to
induce sliding
Flash flood
Severe, short-lived flood events triggered by extreme cloudbursts, glacial lake outbursts,
or the failure of artificial dams or dams caused by landslides, debris, ice, or snow; flash
floods can have impacts hundreds of kilometres downstream, although the warning time
available is counted in minutes or, at the most, hours
Landslide hazard A measure of the probability of a landslide occurring at a particular location within a
particular time
Landslide risk
The probable extent of damage if a landslide occurs; it is a function of the hazard
probability and the damage potential
Landslide dam A natural river dam formed by the rock, earth, debris, and/or mud transported by a
landslide; easily formed in the steep, narrow valleys of high rugged mountains and can
fail catastrophically causing major downstream flooding and loss of life
Mass movement
(gravitational mass
movement)
The downhill movement of surface materials under the influence of gravity but assisted by
buoyancy due to rainfall or snow melt; see also landslide
ix
Mitigation measures
Structural and non-structural measures undertaken to limit the adverse impact of natural
hazards, environmental degradation, and technological hazards; sustained actions
taken to reduce or eliminate a long-term risk to people, infrastructure, and property
from hazards and their effects; measures taken in advance of a disaster to decrease or
eliminate its impact on society and the environment
Natural hazard
Natural process or phenomenon that may cause loss of life, injury or other health
impacts, property damage, loss of livelihoods and services, social and economic
disruption, or environmental damage
Non-structural
Actions to reduce the effects of floods using non-physical solutions such as land use
planning, floodplain zoning, forecasting, advance warning systems, and flood insurance
Slide
Mass movement with a distinct surface of rupture separating the slide material from
(sometimes ‘landslide’) more stable underlying material (slope failure); the downward sliding material is usually
relatively dry
x
Soil conservation
The set of management strategies for the prevention of soil being eroded from the earth’s
surface or becoming chemically altered by overuse
Soil erosion
The washing away of soil by currents of water, wind, or snow
Structural measures
Actions to reduce the effects of floods by physical interventions such as retention basins,
embankments, dredging, diversions, dams, levees, floodwalls, elevating buildings, and
flood-proofing
Flash floods are sudden and intense surges of water that rise and fall very rapidly along a river course (see
Box 1). They differ from riverine floods in terms of their rapid onset and decline, high intensity, and unpredictability
as well as their usually more localized impact in hill and
mountain areas rather than the plains. The flood wave moves
at high velocity, often with a high load of debris, and can
Box 1: What makes a flood a flash flood?
thus be extremely destructive, damaging buildings and other
Intense short-lived precipitation or sudden release of
infrastructure; washing away fields, crops, and animals;
a mass of water compounded by:
and causing loss of life. Flash floods often occur in remote
• increased impervious cover
mountain catchments far from rescue services. They are
• increased stream density
among the most common natural hazards in the Hindu Kush
• increased slope
Himalayan (HKH) region, but they are highly unpredictable
and
and leave little time for warning and evacuation, making it
• decreased channel length
difficult to mitigate their effects.
• decreased surface roughness
• decreased vegetation
Major Factors Contributing to Flash
Floods
Flash floods in the Hindu Kush Himalayan region have a variety of causes, the main ones being localized intense
rainfall, rapid melting of snow, and failure of natural or artificial dams or breakout of ice jams on watercourses. A
number of meteorological, geological, and hydrological or hydraulic factors contribute to the development of such
floods. The first is the seasonal (monsoon) rainfall and cloudbursts that occur in the mountain region, compounded
by the high and rapid runoff from areas with steep slopes, an impermeable surface, saturated and/or compacted
top soil, and lack of vegetative cover. Second, landslides and debris flows, which are also common in mountain
areas, can form unstable dams across rivers, leading to the formation of temporary lakes and the development of a
flash flood when the dam collapses. Bridges can also trap debris and obstruct flow, leading to inundation upstream
followed by collapse of the debris accumulation and a flash flood as a result. Similarly, the development of glacial
lakes behind unstable moraine dams poses a threat of a specific type of flash flood – a glacial lake outburst flood
(GLOF) – when the moraine dam collapses (Ives et al. 2010). All of these can be compounded by factors related to
people’s use of the landscape – anthropogenic factors – which increase the intensity and impact of the flood. These
different factors are discussed in more detail in the following paragraphs.
Intense rainfall
Intense rainfall resulting from cloudbursts, monsoon
depressions, or stationary monsoon troughs is a common
cause of flash floods in the Himalayan region. During the
summer, humid winds flow from the Indian Ocean to the
region, which results in intensive orographic rainfall on the
southern aspect of the mountains. Intense rainfall in a small
catchment can lead to a rapid rise in the water level in
streams, especially where watercourses are narrow. Impervious
soil surfaces increase the rate of runoff, and steep slopes result
in acceleration of water velocity and a high rate of discharge
downstream in the form of a flash flood. The flash flood event
in the Kulekhani catchment of central Nepal in July 1993 is
an example of the type of damage that can be caused by flash
floods resulting from intense rainfall (see Box 2).
Box 2: Flash flood in central Nepal, 19–20
July 1993
In July 1993, the Kulekhani catchment in central
Nepal experienced intense rain, which caused
several devastating flash floods that transported
high loads of debris and sediment into the reservoir
of Nepal’s only storage-type hydropower station,
thereby reducing its life by several decades. Other
installations of the power plant were severely
damaged. The highway joining the capital city to
the rest of the country was seriously damaged, with
several bridges and kilometres of paving washed
away. Fourteen hundred lives were lost. The total
impact was so enormous that it pushed the country
back in its development efforts by several years.
Source: Dhital et al. 1993
1
Resource Manual on Flash Flood Risk Management – Module 3: Structural Measures
Rapid snowmelt
Especially in the western Himalayas and Karakoram mountains, rapid melting of snow can have a similar impact to
intense precipitation. These areas experience considerably more winter precipitation than in the east of the region.
Temperatures can rise very rapidly in the early summer leading to rapid melt of snow and high runoff from the bare
rocky slopes. This type of flash flood is well-known in areas like Chitral, where communities may have memories and
knowledge of potential flood paths reaching back through generations (Dekens 2007).
Landslide dam outburst
Landslides and debris flows are very common in the Hindu Kush Himalayan region. Weak geological structures,
active tectonic forces, steep and fragile topography, heavy seasonal rainfall, and river bank erosion are the major
factors responsible. Deposits from landslides and debris flows can block a narrow river course to create a natural
dam resulting in the formation of a temporary reservoir upstream. The water level in the reservoir will rise due to the
continuous inflow from the river. When the water overtops the dam, or its weight exceeds the holding capacity of the
dam, the dam can burst resulting in a sudden torrent of water downstream – a landslide dam outburst flood (LDOF).
Landslide dam outbursts are generally random and cannot be predicted with precision. The Yigong LDOF in 2000
in Tibet Autonomous Region, China (see Box 3), the Tsatichhu LDOF in 2004 in Bhutan, and the Pareechu LDOF in
2005 in Tibet Autonomous Region are notable examples of large LDOF events in the region (Shrestha 2008).
Glacial lake outburst
As in many parts of the world, glaciers in much
of the HKH region are retreating and thinning, a
phenomenon now accelerated by climate change
(Mool et al. 2001; Xu et al. 2007; Eriksson 2009).
This has led to the formation of glacial lakes behind
the end moraines that formed when the glaciers were
at their maximum. The moraine dams are composed
of unconsolidated boulders, gravel, sands, and silt
and are thus structurally fragile and potentially subject
to catastrophic failure. The collapse of a moraine
dam can result in the sudden release of a large
amount of water and debris – a glacial lake outburst
flood (GLOF). The distribution of glacial lakes in
the Hindu Kush Himalayan region and the potential
for outburst is described in some detail in Ives et al.
(2010) and ICIMOD (2011). Though these types of
flash floods are not very frequent in the region, they
threaten populated areas downstream with potentially
serious consequences. The Zhangazanbo GLOF
in China in 1981 is a notable example of a GLOF
that caused significant damage. The flood damaged
the highway and bridges below the lake up to the
Sunkoshi power station, including the Friendship
Bridge between Nepal and China and the diversion
weir of the Sunkoshi hydropower station, and the
erosion and sedimentation led to a marked change in
the landform downstream (Shrestha 2008). The Dig
Tsho GLOF in Bhutan is a further example (see
Box 4).
Box 3: Yigong landslide dam outburst flood, Tibet
Autonomous Region, China, 10 June 2000
One of the most striking examples of a LDOF took place on
the Yigong River in eastern Tibet in 2000. As a result of a
sudden increase in temperature, a huge amount of snow and
ice melted in the region, which led to a massive, complex
landslide on 9 April in the upper part of the Zhamulongba
watershed on the Yigong River, a tributary of the Yarlung
Zangbo River. Within eight minutes, about 300 million cubic
metres of debris, soil, and ice was dumped across the river
bed, forming a landslide dam 100 m high, 1.5 km wide
(along the river), and 2.6 km long (across the river) (Shang
et al. 2003). The blocked river formed a lake behind the
dam; the inflow from the river was about 100 m3 per second
and the lake level rose by about a metre per day. An attempt
was made to dig a large trench to release the water but this
failed. On 10 June 2000, the dam broke leading to a huge
flash flood downstream. The flood was 1.26 × 105 m3/s,
with a maximum depth of 57 m and maximum velocity of
11.0 m/s. The peak flood was 36 times greater than the
normal flood. Tongmai Bridge, the highway between Yigong
Tea Farming Base and Pailong County, and two suspension
bridges in Medong County were destroyed. There were no
injuries or deaths on Chinese territory, but on the Indian
side of the border, the damage was on a scale seldom
seen and resulted in the deaths of 30 people, with more
than 100 missing. The flood in the Brahmaputra River as
it entered India was 1.35 × 105 m3/s (Zhu and Li 2000;
Zhu et al. 2003). More than 50,000 people in five districts
in Arunachal Pradesh were rendered homeless, and more
than 20 large bridges, lifelines for the people, were washed
away. The total economic loss was estimated at more than
USD 22.9 million.
Source: Shrestha 2008
2
Anthropogenic factors
In addition to natural factors, there are a number
of anthropogenic factors that either contribute to
creating the conditions that favour the development
of flash floods or increase the associated risk, in
particular settlements on flood plains, urbanization,
deforestation, and failure to maintain or manage
drainage systems (Shrestha 2008). Many of these
are the direct result of population growth and the
associated pressure on natural resources and land
for food production and development. For example,
deforestation and soil erosion in mountain and hill
regions as a result of increased farming on marginal
lands, demand for fuel, and poor construction of
roads and trails, means an increase in landslides and
slips and the associated potential for the creation
of landslide dams. Similarly, poor management of
watersheds, introduction of intensive agricultural
practices, and poor land use practices leading to
deforestation, degradation, and soil compaction or
loss (see, for example, Mehta 2007; Dixit 2003), as
well as building footprints that reduce infiltration and
failure to maintain drainage systems, can all lead to
decreased retention of precipitation and higher runoff,
directly contributing to flash floods. Poorly planned
infrastructure construction, such as the construction
of bridges (especially low truss or box bridges) across
narrow river channels and altering of water channels,
can also contribute to the development of flash floods
by hindering the free flow of water and debris and
causing blockages which can burst out. The failure of
man-made structures such as dams, embankments,
and water reservoirs can also be a direct cause of flash
floods (see Box 5).
Box 4: Dig Tsho glacial lake outburst flood, Nepal,
4 August 1985
The Dig Tsho GLOF of 4 August 1985 was triggered by an
ice avalanche from the Langmoche Glacier which induced
a dynamic wave on the lake. Vuichard and Zimmerman
(1987) reported that an ice mass of 100,000 to 200,000 m3
dislodged itself from the overhanging glacier tongue and
plunged into the lake causing the moraine dam to breach.
According to the report, the flood began in the early afternoon
and lasted for 4–6 hours. By reconstructing the hydrograph,
Vuichard and Zimmerman estimated that the peak flood had
been 1,600 m3/s; Cenderelli and Wohl (2001) estimated
a higher peak discharge of 2,350 m3/s. Local witnesses
reported that the flood surge moved rather slowly down
the valley as a huge black mass of water and debris. The
mean velocity of the surge front was between 4 and 5 m3/s.
The most significant impact of the GLOF was the complete
destruction of the newly-built hydropower station in Thame,
which had cost an estimated USD 1.5 million.
Source: Bajracharya et al. 2007
Box 5: Collapse of Gouhou Reservoir, China,
27 August 1993
At 23:40 on 27 August 1993, the Gouhou Reservoir dam in
Qinghai Province (total capacity 3.3 × 106 m3) collapsed and
the water in the reservoir flooded out. The maximum flow rate
at the dam was 3,000 m3/s. Some 2.68 × 106 m3 of water
was released. The flood rushed into Qiabuqia Town, 13 km
downstream, with a population of about 30,000. More than
1,000 houses were destroyed instantly, 288 people died, and
more than 1,000 were injured. The direct economic loss was
around USD 27.7 million (RMB 160 million). The collapse was
not due to any extraordinary storm or flood but took place in
fair weather.
Source: Zhou and Wan 2005
In addition to factors that increase the likelihood of
flash floods occurring, the associated risk is also increasing as a result of, for example, increased urbanization in
mountain areas; location of settlements, roads, and infrastructure close to narrow water courses and on floodplains;
and lack of awareness of migrant populations of local risk factors.
Flood Risk Management
Flood risk is a combination of the probability of a flood occurring (the hazard) and of the potential adverse
consequences of the flood for human health, the environment, cultural heritage, and economic activity (the
vulnerability) (FRM 2009). The combination of these factors gives a measure of the risk, the likelihood and likely
cost of damage occurring to people and property. Thus flood risk management has two major aspects: changing
the conditions so that a flood is less likely to occur, and reducing the vulnerability of people and property if a flood
does occur. When considering measures to reduce risk, it is necessary to achieve a balance between the economic,
social, and environmental dimensions (Klijn et al. 2009).
Measures to reduce flood risk can be divided into two categories: non-structural and structural. A complete plan for
flood risk management usually contains elements of both approaches.
3
Non-structural measures
Non-structural measures refer to any measure that does not involve physical construction but instead uses
knowledge, practices, and/or agreements to reduce the potential impacts of a flood. Non-structural approaches
can be a cost-effective alternative to traditional engineering solutions. Typical approaches include policies and laws,
raising public awareness, and training and education. Such measures offer a variety of possibilities including the
installation of early warning systems, soil management and acquisition policies, insurance, perception, awareness,
and public information actions, emergency systems, and post-catastrophe recovery, all of which can help mitigate
flood-related problems. Non-structural measures are generally more sustainable and less expensive than structural
measures. They can be the most effective way of managing flash floods, which are often localized and difficult to
predict, in contrast to riverine floods. However, non-structural measures are only efficient with the participation of
a responsive population and an organized institutional network. Non-structural measures for management of flash
flood risk are discussed in detail in Module 2 of this Resource Manual (Shrestha 2008).
Structural measures
Structural measures refer to any physical construction designed to intervene, control, or mitigate the potential
impacts of floods. In other words, they refer to the use of engineering techniques to achieve hazard resistant and
resilient structures or systems. They include measures to increase infiltration and reduce rates of runoff in upper
catchment areas, measures to stabilize slopes and reduce the likelihood of landslides and mudflows, and structures
designed to keep floodwaters away from people and property, such as dams, levees, diversions, and check dams.
Structural measures for flash flood management can be separated into four broad groups in terms of the overall
focus (APFM 2007): activities in the whole of a catchment area, activities to shape retention, regulating rivers and
streams, and river conservation. Activities in the whole of a catchment area focus on measures to limit the speed of
surface runoff and limit flood erosion, including the promotion of good farming practices, terracing, and stabilizing
slopes and drainage ditches. Water retention activities can help reduce the flood wave by reducing the amount of
water that runs off a catchment surface. Activities to promote retention include constructing small reservoirs, dry
reservoirs, and polders and building small dikes and dams. Rivers and streams can be regulated by measures to
slow the speed of water flow, including reducing the slope of the riverbed to check erosion, constructing barriers and
thresholds, and building different types of anti-debris dams, dikes, and embankments. Finally, in the river valley, the
river corridor can be shaped by controlling the depth and slope of the river bed so that flood water is directed away
from high impact areas.
Bioengineering is a way of using living plant material to provide reinforcement and create guiding structures.
Bioengineering methods are not structural methods in the strict sense of the term as structural methods are usually
considered to refer to physical or artificial construction. However, bioengineering techniques are closely related to
the physical structural approach, and are often used in combination with and complementary to structural measures.
Thus bioengineering methods have also been included in this volume on structural measures.
The Resource Manual
The present volume focuses specifically on the structural measures that can be used to manage risk from flash
floods. In order to design appropriate structural measures, it is first necessary to understand the mechanisms
of the major contributing factors so that these can be addressed. Thus the next two chapters are devoted to an
understanding of the development and characteristics of landslides and debris flows (Chapter 2) and of landslide
dam lakes and glacial lakes (Chapter 3). These, together with intense runoff, are the major causes of flash
floods in the Hindu Kush Himalayan mountains. The following three chapters describe various techniques for
structural measures that can be used to reduce flash flood risk. These have been divided into three major groups:
bioengineering or vegetative measures (Chapter 4); physical measures for slope stabilization and erosion control
(Chapter 5); and river training structures (Chapter 6). The final chapter (Chapter 7) looks at the whole approach of
integrated water resource (and flood) management in terms of implementing structural measures.
4
Chapter 2: Landslides and
Debris Flows
Landslides and debris flows are among the major causes of flash floods in the Himalayan region as the rock, earth,
debris, and mud that they transport can be deposited across river beds to form an unstable barrier to river flow – a
landslide dam – which may fail catastrophically leading to sudden downstream flooding. It is easy for landslide dam
lakes to form in the steep and narrow valleys found in high rugged mountain areas because only a relatively small
amount of material is needed to block such valleys (ICIMOD 1991a). Thus one way of mitigating flash floods is to
take measures to reduce the occurrence of landslides and debris flows. For this, it is first necessary to understand
how they form.
What is a Landslide?
A landslide is defined as the movement of a mass of rock, earth, or debris down a slope (Cruden 1991). A debris
flow is effectively a type of landslide in which the mass is saturated or oversaturated with water, thus forming a slurry
which flows down the slope. A debris flow is actually an intermediate form between a landslide and a sediment
laden flood, with the characteristics of mixed loose solid material (Wu and Li 2001). Debris flows are differentiated
from floods by the higher unit weight of flow (>1.3 tonnes/m3) and gradient (>1%).
Classification of landslides
Landslides are classified in many different ways. Varnes’ (1978) classification is one of the most commonly used; it is
based mainly on the nature of the source materials and the type of movement involved. The main types of landslide
are summarized in Table 1, illustrated in Figure 1, and described briefly in the following section.
Fall. A fall is a sudden movement of material that detaches from a steep slope or cliff. It occurs as a single free fall
or a series of leaps and bounds down the slope. Depending on the material involved, it can be a rock fall, soil fall,
debris fall, earth fall, boulder fall, or other fall. Falls take place on slopes of 45° to 90°.
Topple. A topple occurs when a block of rock tilts and rotates forward on a pivot or hinge point and separates from
the main mass or surface, falling to the slope below and subsequently bouncing or rolling down the slope. A topple
can be triggered by gravity, water, ice, or wind, and the rate of topple can vary from very slow to extremely rapid.
Table 1: Types of landslide (abbreviated classification of slope movements)
Type of movement
Type of material
Engineering soil
Bedrock
Predominantly fine
Predominantly coarse
Fall
Earth fall
Debris fall
Rock fall
Topple
Earth topple
Debris topple
Rock topple
Slide (rotational/translational)
Earth slide
Debris slide
Rock slide
Lateral spread
Earth spread
Debris spread
Rock spread
Flow
Earth flow
Debris flow
Rock flow
Complex
Combination of two or more principal types of movement
Source: abbreviated version of Varnes’ classification (Varnes 1978)
5
Figure 1: Types of landslide
a. Fall
c. Slide
b. Topple
Talus
d. Spread
e. Flow
Original position
Moving mass
Slide. A slide is a down slope movement of rocks or soil occurring mainly on surfaces of rupture or relatively thin
zones of intense shear strain. There are two major types of slide: rotational and translational. Rotational slides occur
most frequently on the surface of homogeneous slopes, whereas translational slides take place on more or less
plane surfaces.
Spread. A spread is an extension of a cohesive soil or rock mass combined with a general subsidence of the
fractured mass of cohesive material into softer underlying material (Cruden and Varnes 1996). It occurs when a
large block of soil spreads out horizontally after fracturing from the original base. Spreads usually occur on gentle
slopes of less than 6°.
Flow. A flow is in effect a liquefied landslide containing a high proportion of unconsolidated, saturated material
suspended in water. Flows can carry material ranging in size from clay to boulders, and may contain woody and
plant debris such as logs and tree stumps. Flows can be of different types according to the materials they contain
and the speed of flow, including mudflows, earth flows, debris flows, creeps, and debris avalanches. The rate of
flow may vary from slow to very rapid (from 5 km/hr to 40–50 km/hr or more in extreme cases) and the flow can
cover several kilometres. The flow starts when unconsolidated material becomes saturated and can be triggered
by intense rainfall, glacial melt, and earthquakes and precipitated by shallow landslides or in some cases by the
collapse of a riverbed. Debris flows are relatively common in steep mountain areas with high rainfall. They are
extremely destructive of life and property because of the mass and high speed of flow (Takahashi 2007). They can
cause significant erosion of the substrates over which they flow, thereby increasing their sediment charge and further
increasing their erosive capabilities (Nettleton et al. 2005).
Complex movement. A complex movement is a combined process of two or more of the types of movement
mentioned above. Large-scale movements are usually complex, as in a rock/debris avalanche, for example.
6
Figure 2: Forces acting on a slope
Mechanism of Landslide Formation
A landslide is a by-product of slope instability. Slope
stability is the state of balance between driving forces
and resisting forces acting on the earth’s surface
(Figure 2). The driving force tends to pull materials
down a slope whereas the resisting force holds the
material where it is. When the driving force exceeds
the resisting force, the slope becomes unstable and
the materials on the slope will start to move. This
can result from an increase in the driving force, a
decrease in the resisting force, or a combination of
the two.
Driving
force
Resisting force
along a potential
failure surface
Driving force (shear stress)
The driving forces on a slope make up the shear
stress or driving stress (t). The main force contributing to shear stress is gravity. The slope angle, type of material,
water content, earthquakes, and anthropogenic activities can all contribute to the effect of gravity. The driving force
increases with increasing steepness.
Resisting force (shear strength)
The resisting forces on a slope make up the shear strength (S). These forces develop as a result of the internal
friction caused by interlocking of molecular particles. The resisting forces depend upon the shear strength of
the slope materials, which is a function of cohesion and internal friction (shear strength = cohesion + internal
friction). Cohesion (C) is the innate ‘stickiness’ of a material: the strength of attraction or bonding of molecules. For
example, clay and granite are both cohesive whereas dry sand is non-cohesive. Internal friction refers to the friction
between grains within a material and is expressed in terms of the coefficient of friction or angle of internal friction
(f). It depends on how slick the grains are and is a function of the type of material and how strongly the grains are
pressed together by gravity, which is expressed in terms of the slope normal component of gravity or normal stress
(s). Water plays an important role in reducing the resisting force.
The forces acting on a point along the
potential failure plane are illustrated in
Figure 3.
Figure 3: Force diagram for thin to thick translational slides
Factor of safety
FS =
Shear strength (resisting force)
Shear stress (driving force)
.
FS < 1 indicates unstable slope
conditions, FS = 1 indicates that the
slope is at the point of failure, and
G
σ

The factor of safety, FS, for a rock
slope is the ratio of the total force
resisting sliding (resisting forces) to the
total force tending to induce sliding
(driving forces) (Li et al. 2001) along
an assumed or known rupture surface.
In other words, it is the ratio between
the shear strength and the shear stress:
τ
ter
Wa
lure
fai
tial
oten
P
e
tabl
S =
s =
m =
t =
G=
e
plan
S
Shear strength
Normal stress
Pore water pressure
Shear stress
Slope normal component of gravity
µ
7
FS > 1 indicates stable slope conditions. According to Vazirani and Ratwani (1980) the factor of safety against
sliding should be not less than 1.5.
In order to assess the potential of failure on a slope, it is important to calculate the factor of safety.
The equation for the factor of safety is given by
FS =
C + (γ
mγ w ) z cos β tan φ
γ z sin β
where
C = the cohesion of the material,
g = the unit weight of the material,
m = the proportion of the slab thickness that is saturated,
gw = the unit weight of water,
z = the thickness of the slope material above the slide plane,
b = the slope of the ground surface in degrees assumed to be parallel to the failure plane, and
f = the internal angle of friction in degrees.
The model relies on several simplifying assumptions. The values take into account factors such as the level of the
water table, the depth of the failure surface, and soil weight but do not account for the impact of adjacent factors
such as upslope development or downslope modifications of the hill slope or accentuating factors such as ground
vibrations or acceleration due to earthquakes.
The shearing resistance of a soil can also be determined in the laboratory using the direct shear test, triaxial shear
test, unconfined compression test, or Vane shear test.
Factors Contributing to the Development of Landslides
A landslide is the result of slope failure. The causative and triggering factors include anything that leads to an
increase in the shear stress (driving force) on the slope, a decrease in the shear strength (resisting force), or a
combination of the two. The factors can be geological, morphological, physical, and/or human (Cruden and Varnes
1996). They are generally divided into natural factors and man-made (anthropogenic) factors. The major natural
factors include high relief and steep slopes, soil type, vegetation cover, river bank erosion, heavy precipitation,
earthquakes, folding and faulting, and weathering. In particular, unconsolidated deposits, strongly weathered and
fractured rocks, and earth become saturated during heavy rainstorms and are then very prone to slide on a steep
slope. The major anthropogenic factors are inappropriate land use and poor watershed management practices such
as deforestation, extension of agriculture on steep slopes, intensive agriculture or unsuitable crops, and overgrazing;
poor water management; unplanned settlements; and poorly planned construction of roads, trails, and other
infrastructure.
Debris flows are characterized by high water content. The causative factors for debris flows include the above, but of
these the most important are steep slope, loose rock and soil materials, clay minerals, saturated soils, and especially
rainfall or snowmelt generated runoff of sufficient intensity and duration to initiate slope movements (Brooks et al.
2005; Tognacca et al. 2000, cited in Brien et al. 2008).
Mitigation Measures
The first approach to mitigating flash floods is to reduce the likelihood of occurrence of landslides and debris flows,
and thus the creation of landslide dams. This means taking measures to reduce the likelihood of slope failure. Slope
failure mitigation measures can be divided into three categories: control, restraint, and other (JICA 2006). The most
common measures are summarized in Table 2 and discussed briefly below. Specific techniques are described in
detail in Chapter 4 (Bioengineering Measures ) and Chapter 5 (Physical Methods).
8
Control measures
Control measures are applied to prevent or reduce the occurrence of landslides by modifying the natural conditions
that contribute to their formation such as topography, groundwater, surface water, and other conditions. These
measures can be implemented gradually over a long period. The major types of control measure are drainage,
slope protection, and reforming the slope through soil removal.
Increasing the efficiency of surface water drainage can help to limit the infiltration of water into the soil and thus
reduce the landslide potential. The two main approaches are drainage collection and drainage connection to
remove the collected water. It can also be useful to limit water entering the ground from water bodies such as
ponds, irrigation channels, and paddy fields by making the banks and bottom of these water bodies water resistant
or impermeable with materials such as plastic sheeting, clay, asphalt, and concrete.
Groundwater control measures are designed to remove groundwater from the slope and prevent more water flowing
into the mass from outside. This can mean draining surface water and shallow sub-surface water up to 3 m below
the surface as well as draining water from deeper layers by boring. Where the slip surface is in the shape of a
valley, drainage wells can be dug at the bottom of the valley and water can be drained out to the wells through a
borehole. A drainage tunnel can be built to remove deep underground water, where a large volume of underground
water flows near the slip surface.
Slope protection measures involve using vegetation (bioengineering) or artificial structures to stabilize the slope
and cover the surface to reduce infiltration and increase cohesion. Bioengineering methods are one of the main
approaches used to mitigate debris flows, as minimizing runoff above the slope and binding loose materials on the
slope through a network of roots have the highest priority. Bare slopes without vegetation are more prone to debris
flows than slopes covered by vegetation (Li and Clarke 2007).
The slope can also be reformed to maintain the balance of the landmass, most commonly by removing the sliding
mass from the upper part, sometimes replacing it at the foot of the slope. Such measures require skilled engineering
and can sometimes trigger a landslide, so careful investigation and planning is required before implementation.
Table 2: Mitigation measures
Purpose
Drainage
Surface drainage/groundwater drainage
Control
Slope protection with vegetation
Slope protection using Spraying
structures
Bioengineering
Spraying mortar
Spraying concrete
Restraint
Other
Plastering
Plastering with stones or blocks
Spraying concrete
Crib work
Pre-cast crib wall construction
Cast-in-place crib work
Removal of unstable soil mass
Improvement of slope shape
Excavation
Excavation
Construction of retaining walls
Stacked blocks and stone masonry
Concrete leaning walls
Concrete gravity walls
Concrete crib retaining walls
Pile construction
Anchoring
Counterweight filling
Protection methods
Piling
Anchoring
Counterweight filling
Catch walls
Rock fall prevention
Rock fall prevention measures
Rock fall check measures
Avalanche control
Other construction
Avalanche trapping measures
Fencing
Installing gabions
Temporary protection measures
Source: DWIDP/JICA 2004b
9
Restraint measures
Restraint measures are structural ways of reducing the likelihood of landslides. The most common measures are
construction of retaining walls, anchoring, and pile construction. These techniques have high construction costs and
require specific technical skills; careful investigation and understanding is needed before designing such measures.
Retaining walls can be installed to hinder the small or secondary landslides that often occur at the toe of larger
landslides. Retaining walls are effective against small landslides but cannot resist direct landslide driving forces.
They are usually installed in combination with other measures such as bases of embankment structures and walls for
drainage of horizontal borings. Different types of walls are discussed in more detail in Chapter 5.
Anchors use the tensile force of anchor bodies to stabilize the slope. Steel wires or bars are inserted throughout
the sliding landmass and cemented into the bedrock in order to join the sliding landmass and bedrock together.
Anchor work can be used on moderate to steep slopes. The advantage of anchors is that a large restraint force can
be obtained from a relatively small element. Large restraint forces can be obtained by increasing the number of
anchors.
Steel piles can be used in a similar way as anchors to tie the sliding landmass to the bedrock below in order to
restrain movement. Steel piles of 200–600 mm diameter are designed to resist both shearing and bending stress.
Piles are generally filled with concrete.
The lowest, or toe, area of a landside can be prevented from moving by piling up extra material, and slope stability
can be increased by reducing erosion at the toe. This method is widely implemented because of its reliability and
immediate effect and is sometimes combined with soil removal from the upper part.
Other measures
Other protection measures include installing catch walls on lower slopes to slow landslides, measures to prevent
rock falls that might trigger a landslide, and various types of fencing and other measures to protect installations from
the impact of landslides. These measures also include sabo dams, which are designed to reduce the load carried
downstream by a debris flow. Sabo dams are discussed in more detail in Chapter 5.
10
Chapter 3: Landslide Dam Lakes
and Glacial Lakes
The triggers of flash floods in the Hindu Kush Himalayas include catastrophic failure of landslide dams that retain
landslide dam lakes, and of moraine or ice dams that retain glacial lakes. Flash flooding caused by landslide dam
failure is a significant hazard in the region and is particularly common in the high rugged mountain areas of China,
India, Nepal, and Pakistan (Zhu and Li 2000). The previous chapter looked at ways of preventing the formation of
landslide dams by reducing the prevalence of landslides and debris flows. Another important approach to mitigating
flash floods is to reduce the likelihood of dam failure, and/or to put in place warning and avoidance mechanisms to
reduce the risk in the case of failure. In order to design appropriate measures, it is important first to understand the
formation process, failure mechanism, and risk mitigation techniques for both types of lake.
Landslide Dam Lakes
Formation
Landslide dam lakes can be created as a result of a broad range of mass movements in different geomorphological
settings. Dams form most frequently as a result of rock and earth slumps and slides, debris and mudflows, and rock
and debris avalanches in areas where narrow river valleys are bordered by steep and rugged mountain slopes
(Zhu and Li 2000). A lake then forms behind the dam as a result of the continuous inflow of water from the river
(Figure 4). Only a small amount of material is needed to form a dam in a narrow valley, and even a small mass
movement can be sufficient. This type of setting is common in geologically active areas such as the Hindu Kush
Himalayas where earthquakes occur and slopes are steep. These areas contain abundant landslide source materials
such as sheared and fractured bedrock materials. Large landslide dams are often caused by complex landslides that
start as slumps and transform into rock or debris flows or avalanches. The lake formed along the Hunza River by a
massive landslide in Attabad, Pakistan, in January 2010 is one of the more striking recent examples of a landslide
dam lake in the region.
Landslide dam outburst floods (LDOF) and modes of dam failure
A dam failure is a complex hydrologic, hydraulic, and geological phenomenon which is controlled primarily by the
failure mechanism and characteristics and properties of the dam (Uhlir 1998). In many cases, the breach occurs as
a result of fluvial erosion of the debris or landslide material, with erosion starting at the toe of the dam and moving
progressively upstream to the lake. Piping (the presence of sloping tube-like features in the soil often marked by
seepage) can also lead to internal structural failure of the dam. However, the most common cause of failure is
overtopping by the lake water. Sometimes, piping and undermining of the dam can cause partial collapse followed
by overtopping and breaching. A landslide dam with steep upstream and downstream faces and high pore-water
pressure is also susceptible to slope failure. If the dam has a narrow cross-section or the slope failure is progressive,
the crest of the dam can fail, again leading to overtopping and breaching. Sometimes, lateral erosion of the dam by
the stream or river can cause failure.
LDOF risk mitigation measures
Both structural and non-structural measures can be used to reduce the risk associated with landslide dam failure.
Non-structural measures (discussed in Module 2) include estimation of downstream flooding, estimation of past
floods, early warning systems, and so on.
11
Figure 4: Formation of a landslide dam lake
One of the most efficient structural measures is to lower the water level of the lake. Construction of spillways is the
simplest approach. Pipes, tunnels, outlets, and diversions can also be used to control discharge (Zhu and Li 2000)
and help drain the water. For example, Tangjianshan Lake, formed as a result of a landslide precipitated by the
12 May 2008 Wenchuan earthquake in China, was successfully drained by construction of a sluiceway 13 m
wide, 10 m deep, and 475 m long (Zhuang 2010). In some cases, extensive blasting measures have been used
to excavate new river channels through landslide dams (Zhu and Li 2000). However, it is not always possible to
construct a drainage path, the unstable nature of the dam material can lead to collapse of the channel, and, in the
worst cases, the channel construction itself can destabilize the dam and cause it to breach catastrophically.
Glacial Lakes
Formation
The ways in which glacial lakes can form and their distribution in the Hindu Kush Himalayan region are described
in some detail in Ives et al. (2010) and ICIMOD (2011). They can be divided into two broad categories on the
basis of the damming materials: moraine-dammed glacial lakes and ice-dammed glacial lakes. Moraine-dammed
lakes usually form between the glacier terminus and the exposed end moraine as a glacier tongue thins and retreats
(Figure 5). Ice-dammed lakes can form on the surface of a glacier and between a glacier and valley wall. Lakes
can also form below or within glaciers but these are difficult to identify using normal mapping approaches and in
practice no mitigation measures are possible.
Glacial lake outburst floods (GLOF) and modes of dam failure
Moraine dams are composed of unconsolidated boulders, gravel, silt, and sand and are thus often weak and
susceptible to collapse. As glaciers retreat, the volume (and thus weight) of the water in the lake increases,
12
Figure 5: Formation of glacial lakes: lateral moraine dammed lake with ice (left) and end moraine dammed lake (right)
Gla
ci
er
L a ke
i
ora
ne
ac
l
M
r
Mora
cie
G
ine
Gla
Lak e
L a ke
i
er
Natural
ou t le t
increasing the pressure on the dam. Moraine dammed lakes generally breach as a result of overtopping or piping,
or a combination of the two (Mool 1995). Overtopping can be precipitated by a fall of ice (or rock) into the lake
water, or by a sudden inflow of water from an outburst of a small lake higher up.
Ice-dammed lakes drain underneath the ice. On occasion a pocket of water can burst within the ice. Water within
the glacier’s internal plumbing may build up pressure to such a point that the hydrostatic pressure exceeds the
constraining cryostatic pressure and the water is able to rupture through the glacier ice. Internal water pockets may
burst into existing sub-glacier draining, increasing the discharge to produce a characteristic flow (Sinha 1998).
In 1994, a catastrophic flood from the moraine dammed Luggye Tsho glacial lake in northern Bhutan produced a
flood wave that was more than 2 m high at 200 km from the source (Richardson and Reynolds 2000).
GLOF risk mitigation measures
As with landslide dams, both structural and non-structural measures can be used to reduce the risk associated with
moraine dam failure. The biggest difference is that glacial lakes can be identified and the risk assessed in a planned
way, as the glaciers and moraines have a fixed position, unlike landslide dams which are unpredictable. Thus more
time is available to plan and implement measures to reduce the likelihood of moraine dam failure. The approaches
to and difficulties of GLOF risk assessment in the Hindu Kush Himalayan region are discussed in Ives et al. (2010).
Non-structural measures are discussed in Module 2 and Shrestha et al. (2011). They include risk assessment;
planning measures that reduce exposure and vulnerability; awareness raising and flood preparedness training; and
lake monitoring, early warning, and evacuation.
Structural measures aim to reduce the volume of water in the lake in order to reduce the potential peak discharge
as well as the hydrostatic pressure on the dam. The four main approaches – controlled breaching of the moraine
13
dam, construction of an outlet control structure, pumping or siphoning the water from the lake, and tunnelling under
the moraine or ice dam (Ives et al. 2010; Kattelmann and Watanabe 1996) – are described briefly below. The
challenge with all the approaches is that the moraine dam is inherently unstable, and disturbance of the dam during
construction could itself precipitate collapse. At the same time, construction at the altitudes where glacial lakes
are found is challenging, as is transportation of the appropriate equipment to these often very remote and poorly
accessible sites. A high degree of technical and engineering skill is required, and investigation and planning must be
meticulous. For this reason, structural mitigation is only an option for a very small number of lakes.
Blasting. If glacial lakes are at an early stage in formation and the water volume is small, it may be possible to blast
the dam. However, careful precautions must be taken, in particular, total evacuation of the downstream area before
blasting.
Construction of a spillway or open channel. Construction of a spillway or open channel is one of the most
common techniques used to lower the water level in glacial lakes. However, the risk of rapid erosion of the spillway
and uncontrolled release of water is substantial (Grabs and Hanisch 1993, cited in Kattelmann and Watanabe
1996). An artificial spillway has been used to lower the level of Tsho Rolpa glacial lake in Nepal since 2000. Precise
engineering efforts are needed to implement such types of technique.
Siphoning. Siphoning out the lake water has a
lower risk of inducing catastrophic failure and is
Box 6: Siphoning out Tsho Rolpa glacial lake
more appropriate for remote areas. In theory, lake
The purpose of installing siphon pipes was to test their
levels could be lowered by as much as five metres
performance in the high altitude environment with
with a simple siphon under Himalayan conditions
freezing conditions in winter. The test siphons worked
(Grabs and Hanisch 1993, cited in Kattelmann and
satisfactorily with some maintenance, although the
Watanabe 1996). The number of siphons could be
amount siphoned out was insufficient. The test showed
chosen to give the desired rate of lowering of the
that with sufficient funds, siphoning could be an option
for lowering the water level (Rana et al. 2000).
water level, taking into account the rate of inflow
during the summer monsoon and melt seasons
(Kattelmann and Watanabe 1996). Siphons were
installed successfully in Tsho Rolpa lake in 1995, but
the induced outflow never reached a level that exceeded the inflow, so ultimately the effort was unsuccessful (Box 6).
However, siphoning might still offer a useful option under different conditions.
Drilling and tunnelling. Another possibility is to drill or tunnel through the moraine dam and install a pipe to
drain the water from the lake. Drilling is problematic, however, due to the inhomogeneous and poorly consolidated
structure of the dam material and the risk of piping occurring along the casing installed during drilling (Grabs and
Hanisch 1993). Another option for releasing lake water is to construct a tunnel into the lake from an adjacent
deeper lying valley. Tunnels have been built through solid rock into glacial lakes in Norway and Switzerland
(Kattelmann and Watanabe 1996), but the approach is impractical in the Himalayas with their weak geological
structure and remoteness.
14
Bioengineering is the application of
engineering design and technology to living
systems. In terms of flash flood mitigation,
it refers to the combination of biological,
mechanical, and ecological concepts to
reduce or control erosion, protect soil,
and stabilize slopes using vegetation or a
combination of vegetation and construction
materials (Allen and Leech 1997; Bentrup
and Hoag 1998) (Figure 6).
Figure 6: Bioengineering for soil conservation
Bioengineering techniques used in
combination with civil and social engineering
measures can reduce the overall cost
of landslide mitigation considerably
(Singh 2010). Bioengineering offers an
environmentally friendly and highly cost and
time effective solution to slope instability
Source: DWIDP
problems in mountainous and hilly areas
and is a technique of choice to control soil
erosion, slope failure, landslides, and debris
flows, and thus ultimately to help minimize the occurrence of floods and flash floods.
One of the major differences between physical construction techniques and bioengineering is that physical structures
provide immediate protection, whereas vegetation needs time to reach maximum strength. Thus the combination of
physical and vegetative measures offers a combination of immediate and long-term protection, as well as mitigation
of the ecologically damaging effects of some physical constructions.
Functions of Vegetation
Hydrological functions
Plants play a significant role in the hydrological cycle. Particularly riparian vegetation influences hydrological
processes through effects on runoff; control of uptake, storage, and return of water to the atmosphere; and water
quality (Tabacchi et al. 2000). The hydrological functions of vegetation can be summarized as follows:
Interception: The vegetation canopy intercepts raindrops and reduces their size and mechanical strength, thus
protecting the soil from erosion caused by rain splash.
Restraint: The dense network of coarse and fine roots physically binds and restrains soil particles in the ground,
while the above ground portions filter sediment out of runoff.
Absorption: Roots absorb surface water and underground water thus reducing the saturation level of soil and
the concomitant risk of slope failure.
Infiltration: Plants and their residues help to maintain soil porosity and permeability, thereby increasing retention
and delaying the onset of runoff.
Evapotranspiration: Vegetation transpires water absorbed through the roots and allows it to evaporate into the
air at the plant surface.
Surface runoff reduction: Stems and roots can reduce the velocity of surface runoff by increasing surface
roughness.
15
Stem flow: A portion of rainwater is intercepted by trees and bushes and flows along the branches and stems to
the ground at low velocity. Some rainwater is stored in the canopy and stems.
Engineering functions
Catching: Loose materials have a tendency to roll down a slope because of gravity and erosion, and this can be
controlled by planting vegetation. The stems and roots can catch and hold loose material.
Armouring: Some slopes are very water sensitive. They start moving and/or are easily liquefied when water falls
on them. Vegetation can protect the surface from water infiltration and erosion by rain splash.
Reinforcing: The shear strength of the soil can be increased by planting vegetation. The roots bind the grains of
soil. The level of reinforcement depends on the nature of the roots.
Supporting: Lateral earth pressure causes a lateral and outward movement of slope materials. Large and
mature plants can provide support and prevent movement.
Anchoring: Layers with a tendency to slip over each other can be pinned to each other and the stable
underlying layer by penetration of woody taproots from vegetation which function as anchors.
Draining: Water is the most common triggering factor for slope instability. Surface water drains away more
easily in areas with dense rooted vegetation. Thus draining can be managed by planting small and dense rooted
vegetation such as durva grass.
Choice of appropriate species
In general, it is best to use local species of vegetation in bioengineering as they are already adapted to the growing
conditions, are more likely to be resistant to local diseases, are more readily available, and are likely to be a lower
cost option.
It can also be useful to choose species that can be used for other purposes as they mature, for example, providing
fruit or with branches and leaves that can be used for fuelwood, fodder, or other domestic purposes. This increases
the benefit to local people and their acceptance of the measures.
Major species that can be used for bioengineering purposes in the Hindu Kush Himalayan region include broom
grass (Thysanolaena maxima), Napier grass (Pennisetum purpureum), vetiver grass (Vetiver zinzaniodes), durva grass
(Cynodon dactylon), turf grass (e.g., Festuca arundinacea, Poa pratensis), kans grass (Saccharum spontaneum),
different types of bamboo, giant cane (Arundo donax), Malabar nut (Adhatoda vasica), male fern (Dryopteris filixmas), artemesia (Artemisia spp.), weeping willow (Salix babylonica), mulberry (Morus alba), five-leaved chaste tree
(Vitex negundo), ghogar tree (Garuga pinnata), coral tree (Erythrina variegata), tiger’s milk spruce (Sapium insigne),
and eastern cottonwood (Populus deltoides). Further suitable grass, shrub, tree, and bamboo species can be found
in Singh et al. (1983), APROSC (1991), HMGN (1999), DSCWM (2004), and DSCWM (2005).
Bioengineering in Flash Flood Risk Management
Bioengineering can be used in various ways to reduce flash flood risk. It can be used to stabilize slopes and thus
reduce the risk of landslides and debris flows occurring. It can be used to increase infiltration, to form structures to
temporarily capture and store runoff, and to lower the velocity of runoff, all of which hinder the formation of flash
floods after cloudbursts. And it can be used to change the flow pattern of rivers downstream in order to reduce the
impact of floods that do occur. Bioengineering is often used in combination with structural techniques, either to
reinforce structures or as a complementary approach to increase the overall impact of the measures.
Bioengineering techniques to control slope failure phenomena
Bioengineering can be used to increase slope stability in a variety of ways (Li and Clarke 2007; Lammeranner et al.
2005), in particular
mechanical reinforcement,
controlling erosion,
increasing the infiltration ratio,
16
reducing runoff, and
soil moisture adjustment.
Reinforcement. The dense network of coarse and fine roots from vegetation can work as a reinforcement
mechanism on the slope by binding and stabilizing loose materials. The stabilizing effect of roots is even greater
when roots are able to connect top soil with underlying bedrock, with the root tensile strength acting as an anchor.
Small dense roots also contribute to the shear strength of a slope and thus reduce the risk of landslides and debris
flows. Trees and bamboos can stabilize the whole soil layer in slope terrain, whereas bush and shrub roots mainly
protect soil up to 1 m deep, and grasses can conserve top soil to a depth of around 25 cm (Jha et al. 2000).
Erosion control. Bare soil-covered slopes are easily affected by the splash effect of intense rain leading to heavy
erosion. The surface runoff rate is also very high, and the flowing water can carry the soil particles away and trigger
a debris flow. A dense cover of vegetation protects the soil from splash effects and reduces runoff velocity, while the
roots bind the soil particles, thus hindering surface erosion.
Soil infiltration. As decayed roots shrink, they leave a gap which provides a passage for water seepage, which
leads water away from the surface and reduces the likelihood of surface soil saturation. This reduces slope instability
and hinders the development of debris flows.
Reducing runoff. Vegetation can be used to reduce runoff in a number of ways including trapping of moisture
in leaves and branches, slowing the flow of water across the rough surface, increasing infiltration, and through
structures designed to deflect flow away from the top of a slope and channel it along a desired pathway down the
slope.
Soil moisture adjustment. Soil moisture is a key factor in slope stability. Vegetation can directly influence soil
moisture through interception and evapotranspiration. In interception, precipitation is captured by the vegetation
canopy and returned directly to the atmosphere through evaporation. The rate of interception varies according to
various factors including leaf type and size, canopy density, temperature, and humidity. In evapotranspiration, the
plants channel moisture from the soil to the leaves and stems, from where it returns to the air via evaporation. These
two processes combine to reduce the overall soil moisture content.
Choice of techniques. Different bioengineering techniques are used to control erosion and slope failure in different
parts of the world. The techniques suitable for a particular area should be selected on the basis of availability of
resources, site condition, and required function. Table 3 shows the appropriate bioengineering techniques for
controlling different types of landslide and debris flow hazards. Details of the techniques are given in the latter part
of this chapter.
Bioengineering to reduce the volume and velocity of runoff
High runoff can directly cause development of a flash flood from a small catchment following a heavy localized
rainfall event. The aim of bioengineering in this case is to slow and trap the runoff in order to reduce the rate of
outflow from the catchment. Appropriate techniques include palisades, grassed water ways, brush layering, bamboo
fencing, wattle fencing, and similar measures.
River training
The impact of a flash flood can be reduced by measures designed to direct and reduce the speed of the flood wave
in the river downstream. These measures are a part of so-called ‘river training’ techniques, which are undertaken
to improve a river and its banks in order to change the waterway pattern and reduce the velocity of flow, hinder
erosion, reduce transportation of sediment, and guide flood waves into a less destructive path. The most common
river training measures involve construction of physical structures such as banks and spurs (described in Chapter 6).
However, used alone, these techniques may have a marked negative effect on the environment and landscape,
as well as being expensive. Bioengineering techniques used alone or in combination with physical measures offer
a low-cost approach that is easily implemented by local communities and provides an environmentally friendly
environment for local flora and fauna.
17
Table 3: Basic techniques for bioengineering
Phenomenon Erosion problem and
condition
Suitable bioengineering techniques
Landslide
Smoothing to a suitable slope gradient
Diversion canals, channel lining, catch drains, waterways
Stone pitching and planting of trees, shrubs, and grass slip
Bamboo fencing with live poles, planting and seeding grass
Terracing and planting with bamboo, trees, shrubs, grass
Live peg fence, wild shrubs, live check dams
Contour strips planted with grass, shrubs, and pegs
Fascines, brush layering, and palisades
Planting bamboo with or without a structure
Check dams planted with deep-rooted species (e.g., bamboo, trees)
Deep-rooted landslide
(>3 m depth)
Slumping
Planar sliding
Shear failure
Cut and fill area at deep
and shallow-rooted
landslide (<3 m depth)
Bare and steep slope or
newly exposed surface
Debris flow
Soil Erosion
Cracking zone
Bamboo fencing above zone; zone covered with polythene sheet
Catch drain with vegetation
Head scarp of landslide
or slope failure
Slope excavated to an appropriate gradient and rounded (when high and steep) and
planted with deep-rooted plants (e.g., bamboo, trees)
Bamboo fencing, planting grass, seeding, and mulching
Jute netting or straw mat covering soil, seeds, and compost mixture; turfing
Stone pitching; planting of trees, shrubs, and grass slip
Planting grass slip and seeding grass
Sediment production
zone
As for landslides
Sediment transportation
zone
Series of gabion check dams, retaining wall, and side wall planted with deep-rooted
species (e.g., bamboo, trees)
Bamboo fencing; grass planting, seeding, and mulching
Sediment deposition
zone
Diversion canal, channel lining, retaining wall, and side wall planted with trees, shrubs,
and grasses
Plantation of deep-rooted species (e.g., bamboo, trees)
Sheet and rill erosion
Planting of bamboo, trees, shrubs, and grass with or without terracing
Live peg fence, wild shrubs, and live check dams
Contour strips planted with grass, shrubs, trees, and pegs
Fascines, brush layering, and palisades with wild and thorny shrub species.
Gully erosion
Diversion canals, channel lining, catch drains, waterways, cascade retaining wall, and
side wall, planted with trees, shrubs, and grasses
Bamboo fencing with live pegs
Planting of bamboo, trees, with or without check dams
Series of retaining walls and plantation
Vegetated stone pitching in small gullies and rill beds
Erosion on bare land,
Planting of deep-rooted species (e.g., bamboo, trees)
degraded steep sloped
Bamboo and live peg fencing and live check dams
land, dry and burnt area Vegetated stone pitching in small sheets and rill beds
Degraded shifting
cultivation areas, newly
excavated or exposed
areas on terrace bund,
degraded forest, and
grazing land
Bamboo fencing with live poles, planting and seeding grass
Planting of bamboo, trees, shrubs, and grass with or without terracing and structure
Live peg fencing and live check dams
Vegetated stone pitching in small gullies and rill beds
Contour strips planted with grass, shrubs, trees, and pegs
Planting fascines, brush layering, and palisades
Water induced degraded
land (spring, water
source damaged area,
canal command area)
Planting of bamboo, trees, shrubs, and grass with or without terracing and structure
Planting of deep-rooted species (e.g., bamboo, trees)
Live peg fences and live check dams
Vegetated stone pitching and loose stone masonry walls or check dams
Cut and filled area or
newly exposed area on
slope*
Jute netting and straw mats covering soil, seeds, and compost
Live peg fences and stone masonry walls
Plantation, seeding, and planting grass
Live wattling with terracing and seeding
*Exposed slope surfaces must be carefully maintained. A cut and newly exposed slope surface should usually be covered, depending on
the type of soil material and other factors.
Source: DWIDP/JICA 2004a
18
The use of bioengineering techniques alone is mainly confined to river bank stabilization. By their nature, river banks
provide a good environment for growth of vegetation. Left alone, banks usually have dense vegetation as the river
provides nutrients in the form of silt and water to support growth. If vegetation is sufficient, both on the bank and in
the river bed, it can stabilize the bank, lessen erosion, reduce the speed of flowing water, and reduce scouring by
a flood wave. Where vegetation has been reduced or removed, it can be replaced by carefully selected planting of
appropriate species to achieve the desired effect. Structures formed from a combination of dead and living plant
material can also be used to guide the river course and prevent flood surges entering into settlements and farmland.
The plants can provide additional benefits for the local population like fodder, fruit, and firewood, but this is
secondary to the protective function.
As in slope protection, bioengineering can involve building a structure such as a fence to provide immediate
protection, but using living branches that will take root and become an increasingly strong barrier. It can also
involve a combination of dead and live vegetation, with a framework made of bamboo or timber, intertwined
with living plants to grow and strengthen the structure. A good example is that of a permeable protection wall
constructed out of bamboo porcupines (see river training chapter) intertwined with living plants to form a ‘green
wall’. Placed in the water or in regularly flooded areas of the bank, these structures trap silt in which they slowly
become embedded, creating a strong stable self-sustaining bank with bamboo reinforcement held together by roots
and vegetation.
In general, bioengineering is used in combination with physical techniques in river training rather than on its
own. It is highly recommended as a means of reducing the impact of physical measures on the local ecology and
landscape, and also for providing long-term strengthening of structures such as embankments.
Common Bioengineering Techniques in the Hindu Kush Himalayas
The selection of the appropriate bioengineering treatment for a particular area depends on the site conditions,
and requirements. Resource availability is a crucial factor. The following sections describe some of the techniques
that can be used to control soil erosion, debris flows, landslides, and floods and flash floods in the Hindu Kush
Himalayan region.
Bamboo fencing
Bamboo fencing can be used to prevent soil creep or surface erosion on a slope (Figures 7 and 8), to hinder gully
extension, particularly in seasonal water channels, and to control flood waves along a river bank. Live bamboo pegs
can be used for the main posts so that the whole structure becomes rooted. The growing bamboo can be further
interleaved between the posts (as in a wattle fence) to increase the strength of the fence. Shrubs and grasses are
planted on the upper side of the fence to hold small soil particles. The main purpose is to trap loose sediments on
the slope, to improve the conditions for growing vegetation, and to reduce the surface runoff rate.
Figure 7: Sketch of bamboo fencing
Bamboo poles 40 cm apart
1.5 m
45 cm
19
Figure 8: Bamboo fencing on a slope – the posts (pegs) are live bamboo that will sprout to provide foliage (left);
detail showing live bamboo peg with sprouting leaves (right)
Source: DWIDP
Materials
Live bamboo pegs or strong bamboo poles about 1.5 m long and 10–15 cm in diameter
Digging tools
Seeds or plants of grasses or shrubs
Installation
1.
2.
3.
4.
5.
6.
Starting from the base of the slope, mark the line for the fence with string.
Dig a long pit about 45–50 cm deep along the contour of the slope for each line of fencing.
Insert a row of bamboo poles or pegs 40 cm apart into the pit and back fill the pit to stabilize the poles.
Weave split bamboo or branches in and out between the poles to form a semi-solid face.
Plant small grasses and/or shrubs along the upper side of the fence.
Regular maintenance is important to ensure longevity of the fence. Any broken sections should be replaced
immediately.
Brush layering
In brush layering, live cut branches are interspersed between layers of soil to stabilize a slope against shallow
sliding or erosion. Fresh green cuttings are layered
in lines across the slope (Figures 9 and 10). As the
roots grow, they anchor and reinforce the upper soil
Figure 9: Brush layering
layers (up to 2 m depth), and the foliage helps to
catch debris (Howell 1999, cited in Lammeranner
et al. 2005). Some toe protection structures such as
a wattle fencing, fiberschine, or rock riprap may be
required to support brush layering.
Materials
Branches of different age and diameter cut from
rooting woody plants of different species
(e.g., willow, alder, populus spp., garuga spp.,
Malabar nut [Justicia adhatoda], mulberry, fiveleaved chaste tree [Vitex negundo]). Branches
should be at least 1 m long and 4 cm in diameter
20
Source: Keshar Man Sthapit
Figure 10: Brush layering
Live branches (dormant)
Immediately after planting
Live branches (after sprouting)
One year after planting
Mixed plants of different easily growing species, both rooted and freshly cut
Shovels or other digging tools
Measuring tape and string line to calculate and mark the surface
Installation
1. Mark lines across the slope to be planted at intervals of 0.5–1.0 m upwards from the base. The slope should
have an inclination of at least of 10–20%. Dig a small channel along the line by hand or machine.
2. Cut fresh branches with a right angle at the top and 45° angle at the bottom. If possible, cut the branches on
the same day that they are to be planted. Ensure branches are at least 1 m long with a mixture of different
species. This will allow the root system to penetrate deeper into the soil, giving greater chances of survival and
producing mixed vegetation.
3. Place branches in the dug terrace, with only ¼–⅕ of their length protruding (Figure 10).
4. Place rooted and unrooted plants of species that grow easily 0.5–1.0 m apart among the layers of branches.
5. Regular supervision and care is needed until the branches are fully rooted.
6. The pre-monsoon season is good for installing brush layering. If the site is moist, installation can be done in any
season.
Brush mattress
A brush mattress is a layer of interlaced live branches placed on a bank face or slope, often with a live fascine
and/or rock at the base (Figures 11 and 12). The aim is to provide a living protective covering to an eroding bank
to hinder erosion, to reduce the river velocity along the bank, and to accumulate sediment. The mattress is generally
constructed from live stakes, fascines, and branches from species that root easily, but can be made from any brushy
and woody branches to provide immediate and effective protection. A layer of biodegradable material such as
loosely woven jute can be placed under the mat on steep slopes to increase stability if the soil is very loose. The
mattress that is formed protects the surface of the bank until the branches can root and native vegetation becomes
established.
Materials

Live branches 2–3 m long and approximately 2.5 cm in diameter
Fascine bundles
Live and/or dead wooden stakes
Digging tools (shovel)
21
Figure 11: Cross-section of brush mattress
Brush mattress after rooting
Brush mattress at the time of installation
Wooden stake
Wooden stake
Live branches
Live
Leafy branches
Live stake
Live
Mean water level
Channel bottom
Mean water level
Channel bottom
Figure 12: Installing brush mattress
Source: © Urban Creeks Council
Installation
1.
2.
3.
4.
Prepare the site by clearing away large debris and other materials.
If desired, cover the slope with a layer of biodegradable material, e.g., jute netting, to provide extra stability.
Dig a horizontal trench 20–30 cm deep at the toe of the bank or slope.
Lie the cuttings flat on the graded slope in an overlapping crisscross pattern with the root ends pushed into the
soil in the trench to below the water level and the growing tips placed at a slight angle in the direction of the
stream flow (if on a stream bank) or parallel to the slope.
5. Branches should be placed at a density of approximately 4 branches every 15 cm.
6. Pound wooden stakes between the branches into the soil to half their length and about 1 m apart.
7. Wrap wire around the stakes and over the branches as tightly as possible.
8. Pound the wooden stakes further in to tighten the wire and press the branches down onto the slope.
9. Push live stakes into the ground between the wooden stakes.
10.Place bundles of fascines along the trench at the base of the slope over the bottom of the branches and cover
with soil, leaving the tops slightly exposed.
11.Fill any voids around and in between the branches with loose soil (from the trench) to promote rooting.
12.Periodic maintenance is required to ensure the mattress is securely tied to the slope.
22
Fiberschine
(adapted from Bentrup and Hoag 1998)
Fiberschine is a roll of material made from coconut fibre used to form a toe protection structure on a slope and
to trap any sediment derived from erosion. The most common use is to stabilize the base of a stream bank or
shoreline, but it can also be used in slope stabilization to support other measures such as brush layering. Live
cuttings from herbaceous plants are planted together with the fiberschine; by the time the fiberschine decomposes,
the vegetation will have stabilized the stream bank or slope. Fiberschine can usually be installed throughout the year,
but the high water season should be avoided along streams. The following describes installation along a stream
bank. The method can be adapted for use on a slope.
Materials

Fiberschine roll
Wedge-shaped wooden stakes 60–90 cm long
Twine or wire
Herbaceous wetland plants or willow twigs
Installation
1. Determine the length of the treatment area and obtain the necessary amount of fiberschine.
2. Place a roll of fiberschine along the toe of the stream bank at the level of the low flow line with approximately
half the roll below the water line and half above. Place additional rolls of fiberschine along the bank for the
extent of the treatment area. Tie the ends of adjacent fiberschine rolls together with strong twine.
3. Secure the fiberschine on both sides with wedge-shaped wooden stakes 60–90 cm) long at 1.5 m intervals.
Cut a 7.5–10 cm deep notch in each stake about 12.5 cm from the top. Secure each pair of stakes together
by binding around the notches. Drive the stakes in so that the twine is secured against the top of the fiberschine
(Figures 13 and 14).
4. Key the ends of the fiberschine into the bank to prevent the flow from entering behind it and protect the ends
with something hard such as rock to prevent scouring.
5. Backfill behind the fiberschine by knocking down the top of the stream bank onto the fiberschine.
6. Plant herbaceous wetland plants or willows into and behind the fiberschine at approximately 15–30 cm
intervals.
Figure 13: Securing the fiberschine
Figure 14: Fiberschine used to reinforce a river bank
Rolled fiberschine
Infill
Bank
Water level
Water level
Twine or wire
Stake
Plants
Notch
Stake
Source: Modified from Bentrup and Hoag et al. 1998
Source: Modified from Bentrup and Hoag et al. 1998
23
Jute netting
Jute netting is a useful way of stabilizing steep slopes of 35–80° where it is difficult to establish vegetation
(Figure 15). Locally available woven jute net is used as a form of armour on the slope and low growing grass is
planted through the holes. The technique is often used in South Asia to reduce landslides along roads. The aim is
to protect the bare slope from rain splash erosion, to improve the condition of the site, and to enable vegetation to
become established by retaining soil moisture and increasing infiltration.
Materials

Woven jute net
Digging tools
Sledgehammer
Live wood pegs
Grass seed or small-rooted tufts of grass
Installation
1.
2.
3.
4.
5.
6.
7.
8.
Trim the slope so that it is even and clear away any hanging masses or depressions.
Spread fertile soil on the bare slope.
Mulch with straw or other soft vegetation.
Start laying netting along a line above the slope to be covered, secure by hammering wooden pegs through the
net at 0.3 m intervals.
Unroll the net down the slope and fix by hammering live wood pegs through it at intervals of 0.5–1.0 m.
Continue until the whole slope is covered by netting.
Sow grass seed or plant small grass clumps through the netting diagonally at a spacing of 10cm by 10cm over
the entire area.
Regular supervision and care is needed until the grass is fully grown.
Figure 15: Jute netting on a cut slope
Source: DWIDP
24
Figure 16: Newly constructed live crib wall made of bamboo (left) and live crib wall on a slope (right)
Source: Madhav Dhakal
Live crib wall
A crib wall is a box structure made of interlocking struts (either logs or precast structures made of concrete, recycled
polymers, or other material) and back-filled with boulders, soil, or similar. They are mainly used to stabilize steep
banks and protect them against undercutting, for example a stream bank or the side of a cutting made for a road,
and are also a useful method for stabilizing the toe of a slope. However, they are only effective where the volume of
soil to be stabilized is relatively small.
In a live crib wall, live branches and well-rooted plants are placed between interlocking logs where they can grow
and develop a root network that further strengthens the wall (Figure 16). If needed, the anchor and cross logs can
be held together with nails or bolts. Vegetated crib walls provide immediate protection, and their effectiveness
increases over time as the vegetation grows. Once the plants become established, the vegetation gradually takes
over the structural functions of the wooden supports (Gray and Sotir 1996 cited in Lammeranner et al. 2005). Crib
walls should be installed at an angle of 10–15° towards the slope to increase stability. Green willow branches can
be used to ensure a quick outcome.
Materials
Live branches 1–6 cm in diameter and long enough to reach from the front to the back of the structure with an
overhang at both ends.
Logs, timber, or bamboo 1–2 m long
Steel reinforcing bars
Excavator or digging tools (shovels), rakes, sledgehammers, knife, measuring tape, level instruments
Rock and soil
Installation
1. Excavate an area 1–2 m wide along the toe of the bank (stream bank or toe of a landslide) to a level around
1 m below the surface and fill with rock (Figure 17).
2. Place a series of large logs on the rock end to end along two lines marking the front and back of the wall.
3. Place smaller logs perpendicular to and towards the ends of the large logs from front to back of the wall to form
the bottom layer of a box-like structure. Allow an overhang of about 15 cm in each direction. The logs can be
fixed together with metal bars and nails.
4. Place a layer of live willow (or other) cuttings from front to back of the wall between the logs, and protruding
over the front logs and extending into the slope behind the back logs.
5. Cover the branches with a layer of rock and soil and press down to fill the box.
(Steps 4 and 5 can also be carried out in reverse order.)
25
Figure 17: Section view of live crib wall installed along a river bank
Live branch cuttings
Erosion
control fabric
Compacted
fill material
Channel
forming flow
Base flow
90–120 cm
Toe of slope
Streambed
Rock fill
60–90 cm
120–180 cm
Source: Adapted from ODNR n.d.a
6. Continue for as many layers as needed to reach the desired height, alternating layers of soil and cuttings and
logs and ending with soil. Each successive course of logs parallel to the bank should be set back by 15–20 cm
from the log beneath.
7. To ensure success, the upstream and downstream sections should be well-secured to the bank to prevent
undercutting.
Live fascines
A fascine is a bundle of sticks or brushwood used in construction, generally to strengthen an earthen structure, fill
ditches, or make a path across uneven or wet terrain. Live fascines are bundles of live branches intended to grow
and produce roots. They can be placed in shallow trenches on a stream bank to reduce erosion across the bank
and increase soil stability (Figure 18). The rooted branches protect the toe of the stream bank from erosion and
improve infiltration. Properly placed, the bundles can also trap debris
and sediment. Live fascines can also be used to reinforce slopes and
increase drainage and infiltration. They are installed perpendicular to
the slope in dug trenches or in existing gullies and rills. The optimum
Figure 18: Installing live fascine
spacing depends on the steepness of the slope, usually 4 m intervals
for slopes of less than 30° and 2 m intervals for slopes of 30–45°.
They are most effective on soft cut slopes or slopes with consolidated
debris. Draining effects can be seen as soon as the fascines are
established (Schiechtl and Stern 1992, cited in Lammeranner et al.
2005).
Materials
Live branches of rooting plants of different species 3–5 cm in
diameter and 50–100 cm or more long
Live wooden stakes ready to sprout 3–6 cm in diameter and
50–100 cm long
Dead wooden stakes 3–6 cm in diameter and 50–75 cm long
Digging tools
Jute or coir string or wire to bind the fascines
26
Installation
1. Prepare the site by clearing away loose material and protrusions and firmly infill any depressions (Figure 19).
2. Mark the lines along which fascines are to be installed. The lines should follow the contours, or be at a desired
angle or along rills and at intervals of 2–4 m up the slope (2 m for slopes of 30°–45°; 4 m for slopes of less
than 30°).
3. Excavate trenches approximately 10 cm deep and 20–40 cm wide along the marked lines starting from the
bottom of the slope and working upwards.
4. Bind 4–8 live branches into bundles (fascines) using string or wire.
5. Place the fascines lengthwise in the trenches.
6. Drive live or dead stakes directly through the fascines every metre or so flush with the top of the fascines, and
where the bundles connect.
7. Drive live stakes into the soil immediately below the fascines, protruding about 7 cm above the fascine.
8. Backfill the trench with moist soil to the side of the fascine, but allow the top of the fascine to show.
9. Riprap (see next chapter) can be used to stabilize the toe of the slope and prevent scouring.
Palisades
A palisade is a fence or wall made from wooden stakes or tree trunks. Palisades were used historically as a defensive
structure. In slope protection, palisades are barriers made from live wood cuttings or bamboo installed across a
slope following the contour in order to trap debris moving down the slope, to armour and reinforce the slope, and
to increase the infiltration rate. Palisades are used to prevent the extension of deep, narrow gullies and the erosion
of V-shaped rills (Figure 20) by forming a strong barrier which stabilizes the gully floor and traps material moving
downwards (Lammeranner et al. 2005). They are also effective on steep landslide or debris slopes. Palisades can be
used on a wide range of sites with slopes of up to about 60°.
Materials

Stakes made from cuttings of rooting plants of different species 3–5 cm in diameter and 30–50 cm long
Cross beam
Gabion wire
Digging tools such as crowbars and shovels
Installation
1. Installation should start from the top of the slope and work down.
2. Clean or trim the site, remove unnecessary irregularities of slope and loose material.
3. Mark the places to be planted. Palisades should be spaced at intervals of about 2 m down a slope of less than
30° and 1 m down a slope of 30–60°.
4. Make holes with the help of a pointed bar or crowbar for planting the cuttings.
5. Trim the top of the cuttings at a right angle to the stem and the bottom at an angle of 45°.
6. Place at least two-thirds of the length of the cutting into the hole and pack the soil around it.
7. Tie the stakes with pieces of gabion wire to a cross beam which is anchored in the sides of the gully and
protected by pegs at either end.
8. On steep gullies and rills, support the palisade by placing a layer of stone and soil in front of and below the
structure (Figure 20).
9. Regular inspection is necessary throughout the year. Broken stakes should be repaired and strengthened to
encourage vegetation to develop.
10. Thinning might be required after a few years.
Wattle fence
A wattle fence is made by weaving flexible branches or vines between posts, rather like a large basket. A live wattle
fence is constructed out of live branches which will root and continue to grow and strengthen the fence. The main
purpose of wattle fences is to catch debris moving down a slope and to reinforce and modify the slope. Different
27
Figure 19: Live fascine installation on a river bank
3. Lightly cover with soil
Live stake
Live or dead stake
2. Lay in fascine
1. Trench
Top of live fascine slightly
exposed after installation
Cross-section
Moist soil backfill
Live fascine bundle
The roots and leaves develop
after installation.
Prepared trench
Stream-forming flow
Base flow
Streambed
Erosion control
fabric and seeding
Geotextile fabric
Toe protection
Source: Adapted from ODNR n.d.b
28
Live stake (60–90 cm spacing
between stout stakes)
Live or dead stout stake
(60–90 cm spacing along bundle)
Figure 20: Palisades
kinds of grass (e.g., broom grass and napier grass) and tree species can also be planted along the fence. Wattle
fences are useful in small shallow short slides as well as for river bank protection if combined with other measures
such as brush layering, live pegs, and rock riprap (Figure 21).
Materials

Sharpened stakes from plants 1 m long and 4–6 cm in diameter
Shorter stakes 0.5 m long and 3–4 cm in diameter
Long and flexible woody cuttings from plants which can root easily
Jute or coir string or wire to bind
Digging tools
Installation
1. Prepare the site by clearing all loose material and protrusions.
2. Mark lines along contours on the slope where the fences are to be installed. Fences should be spaced at
intervals of about 4–5 m down the slope, depending on the site and slope angle.
3. Dig holes at 1 m intervals along the lines for the stakes.
4. Insert 1 m long stakes in the holes and place two 0.5 m long stakes at equal distances between the long stakes.
Both long and short stakes should protrude about
20–30 cm.
5. Dig out a trench at least 15 cm deep along the
contour between the stakes.
Figure 21: Wattle fence used for river bank protection
6. Place the cuttings with their lower ends in the
trench, and bend them down along the line of the
fence. Firm the soil back into the trench. Weave
the cuttings in and out between the stakes one
above another to fill in the fence area.
7. If desired, add soil above the wattle fence for
planting tree and grass seedlings and cuttings.
8. Regular supervision and maintenance is
necessary, including weaving the branches in and
out as they grow.
Source: Sundar Kumar Rai
29
Chapter 5: Physical Methods for Slope
Stabilization and Erosion Control
The bioengineering methods for slope stabilization and erosion control described in the previous chapter have a
number of advantages. They are generally low cost and easy to install, and rather than disintegrating over time,
their strength increases as root systems develop and the structures become more stable. However, such methods
are not usually sufficient to withstand the volume of debris involved in mass failure, and are not appropriate for all
the interventions required to reduce flash flood risk. Physical structures and techniques are also required for slope
stabilization and erosion control. Various types of construction can be used to help retain soil and improve slope
stability. The selection of measures always depends upon the site, the topography, and the required result. Proper
selection and design of any measures plays a very important role in slope stabilization and the control of erosion
and measures should only be undertaken as the result of an integrated planning process. Physical measures are
often combined with bioengineering approaches to obtain the maximum effect.
Some of the major physical methods are described in this chapter. They can be divided broadly into measures to
reduce runoff (terracing, diversions, grassed waterways, conservation ponds), methods to stabilize slopes and reduce
erosion (retaining walls, drop structures, sabo dams), and integrated methods to address specific problems (gully
control, trail improvement), although they all tend to have multiple functions.
Terracing
Terracing is the technique of converting a slope into
a series of horizontal step-like structures (Figure 22)
with the aim of:
controlling the flow of surface runoff by guiding
the runoff across the slope and conveying it to a
suitable outlet at a non-erosive velocity;
reducing soil erosion by trapping the soil on the
terrace; and
creating flat land suitable for cultivation.
Figure 22: A terraced slope in Nepal
Terracing helps prevent the formation of rills,
improves soil fertility through reduced erosion, and
helps water conservation.
Types of terrace
Terraces can be made in a variety of ways. The best
approach depends on many factors including the
steepness of the slope, the intended use, and the soil.
The terraces are constructed with light equipment or
by hand. The spacing between the terraces depends
on the slope of land; the distance between terraces
goes down as the slope increases. The three main
types of terrace are bench, level or contour, and
parallel or channel.
30
Source: Jack Ives
Chapter 5: Physical Methods for Slope Stabilization and Erosion Control
Level or contour terraces are constructed along slope contours with the main aim of retaining water and sediment.
The terrace edge is planted with trees, small plants, and grass, usually with trees on the outward facing edge to
increase stability.
Bench terracing is similar to contour terracing with the difference that the terraces do not strictly follow the contour
line and runoff may run along as well as across the terrace. Bench terraces are primarily constructed to enable crops
to be grown on sloping land, rather than to retain water and sediment. Bench terraces are recommended for slopes
with gradient of up to 33%, but as a result of pressure on land are constructed on slopes up to 50–60% (Sharda et
al. 2007).
Parallel or channel terraces are mainly used in heavy rainfall areas. They are also known as graded terraces as they
have a constant slope or gradient along their length which is used to convey excess runoff at a safe velocity into a
grassed waterway or channel.
Of these three, bench terraces are the most common type found in the mountain and hill areas of the Hindu
Kush Himalayan region. Following is a brief description of bench terraces and a type of contour terracing that is
particularly useful for stabilization. The construction of bench terraces is described in more detail in Box 7.
Bench terraces
Bench terraces are particularly suitable where marked seasonal variations exist in the availability of water. The
approach consists of converting relatively steep land into a series of horizontal steps running across the slope.
These steps can be constructed by simply digging out the clayey soil, or they can be reinforced with locally available
mud, stone, or brick. The terraces help conserve moisture during the long dry season, which is especially important
where there are sandy and loam types of soil, and they help to slow and drain away runoff during the heavy rainfall
monsoon season, which also helps counteract the tendency for sliding. There are three main types (Figure 23):
outward sloping terraces, which are used to reduce a steep slope to a gentle slope;
level terraces, which are used to impound water for paddy cultivation; and
inward sloping terraces, which are the most suitable for steep slopes because they guide the surface runoff
towards the hillside rather than down the slope.
Rainwater can be drained from outward sloping terraces along a ditch constructed along the toe of the riser. In
inward sloping terraces, the riser is kept free from flowing water and is protected by a cover of grass.
Terrace design is influenced by the following factors (Sharda et al. 2007):
soil depth and distribution of the top soil;
slope of land;
amount and distribution of rainfall; and
farming practices and proposed crops to be grown.
When designing the terrace, it is necessary to select the type and determine the desired width, vertical interval and
spacing, length, gradient, and cross-section (Box 7).
Contour terraces
The main aim of contour terraces is to retain water and sediment. Contour terraces are similar to bench terraces,
with the major difference that the terrace is formed along the contour, so that runoff flows across but not along the
terrace. In addition, the terrace edge is planted with trees, small plants, and grass to stabilize it and trap sediment.
The terraces can be constructed by excavating soil from the upper half and using it to fill in the lower half as for
bench terraces, or can be allowed to form naturally using a technique called sloping agricultural land technology
(SALT), or contour hedgerow intercropping (agroforestry) technology (CHIAT).
SALT combines the strengths of terracing with the strengths of natural vegetation to stabilize sloping land and make
it available for farming. Double hedgerows of fast growing perennial nitrogen-fixing tree or shrub species are
31
Figure 23: Types of bench terrace
Outward sloping
Shoulder bund
(optional)
Cutting
Filling
Original slope
Final shape
Level
Original slope
Shoulder bund
Cutting
Filling
Final shape
Inward sloping
Shoulder bund
Cutting
Channel
Filling
Final shape
planted along the contour lines on a slope at a distance of 4–6 metres to create a living barrier that traps sediment
carried downslope by runoff (Tang 1999; Tang and Murray 2004). As the sediment builds up, the sloping land is
gradually transformed to terraced land. The space between the contour hedgerows is used for subsistence and
cash crops. The hedgerows both markedly reduce soil erosion and contribute to improving and/or maintaining soil
fertility through nitrogen fixation at the roots and incorporation of the hedgerow trimmings into the soil. SALT can be
established on farmland slopes with gradients of 5–25% or more.
A combined approach has also been developed for improved terraces in which retaining walls are first constructed
along the contours using cement bags filled with soil supported by bamboo cuttings along the contour. The soil is
then excavated from the upper part of the terraces and used to build up the lower part above and behind the terrace
riser wall to create a level bed; the fertile top soil is kept aside and later spread over the newly terraced fields. Grass
and hedgerow species are then planted on the outermost margins of the terraces above the risers (ICIMOD 2008).
The vegetation improves the terrace stability and increases moisture retention, while the construction means that the
terraces are immediately ready for use, unlike the original SALT technique.
32
Box 7: Design and construction of bench terraces
Step 1: Selection of type
The type is selected according to the rainfall and soil conditions of the area. In general, outward sloping terraces are
constructed in low rainfall areas with permeable soils; level terraces in areas with medium rainfall and/or highly permeable
soils, or for growing rice; and inward sloping terraces in areas of heavy rainfall and less permeable soils.
Step 2: Width
The width of the terraces is determined based on the soil depth, slope, amount and distribution of rainfall, and intended
farming practices. Construction of very wide terraces is more costly, requires deep cutting, and results in a higher riser.
However, at least two metres width is required for ploughing using bullocks (DSCWM 2005).
The formula for calculating the width of the terrace is given by Sharda et al. (2007) as
W=
200 x d
S
where
W = width of the terrace in metres
d = maximum depth of the cut (metres)
S = slope of land (%)
Step 3: Spacing
The spacing is the vertical interval (VI) between two terraces. The terrace spacing depends on the soil type, slope, surface
condition, gradient, depth of cut, and agricultural use. The depth of cut and fill have to be balanced, thus the interval is
equal to double the depth of cut. The depth of cut must not be so deep as to expose the bed rock. The spacing is also
linked to the terrace width.
The soil depth limits the maximum depth of cut and thus the maximum possible vertical interval. At the same time, the width
of the terrace should permit economic agricultural operation. The following steps should be followed to take the different
factors into account (Mal 1999).
• Ascertain the maximum depth of the productive soil by taking soil samples from different locations.
• Decide which crops are to be grown in order to calculate the depth of soil required and thus the maximum possible
depth of cut. The depth of cut should be such that at least a minimum convenient width of terrace is obtained.
• If d is the maximum depth of cut, the vertical drop between two consecutive terraces is 2d = D (Figure 24). And the
corresponding horizontal distance is 100D/S or 200d/S.
• If W is the width of the bench terrace and the riser slope is 1:1, the horizontal distance for a drop D is (W + D).
Figure 24: Design procedure for a bench terrace
Original land
D
W
D
1:1 Riser slope
D/2 (maximum
depth of cut)
33
Therefore,
W + D = 100D/S (Mal 1999).
If the riser slope is 1:1,
VI = D = S x W
100 S
Similarly, for a riser slope of 0.5:1, the horizontal distance for a drop D is (W + D/2).
Therefore,
W + D/2 = 100D/S
or VI = D = 2W x S
200 S
Note: For a given slope, the greater the VI, the greater the width. For a given VI, the steeper the slope, the smaller the
width.
Step 4: Length
The length of the terrace is determined by many factors including the shape and size of the land, degree of dissection of
the land, and permeability and erodibility of the soil. Longer terraces are more efficient for agriculture and cost less to
install, but they may increase the velocity of surface runoff thus increasing erosion (DSCWM 2005).
Step 5: Gradient
It is also important to determine the gradient along the terrace. It would be best to select the gradient in accordance with
the rainfall intensity, soil permeability, and width and length of the terrace. But it can also be determined using the simpler
approach of 1 m for every 100 m terrace length, or not more than 1%. In low rainfall areas with highly permeable soil,
the gradient can be lower than 0.5%, whereas in high rainfall areas with low permeable soil, 1% is preferable to reduce
excess run-off. The gradient can also vary, for example for a terrace around 100 m long, 0.25% for the first 25 m, 0.5%
for the next 50 m, and 1% for the remainder.
Step 6: Cross-section
In bench terraces, soil excavated from the upper half is deposited on the lower half. In other words, the volume of cut
is equal to the volume of fill (Mal 1999). The volume of soil after excavation is calculated in terms of loose cubic metres
(LCM), and is more than the volume before excavation (banked cubic metres or BCM). All calculations should be done in
BCM. Filling in LCM should be calculated such that the proper height will be obtained after settling. For land with a slope
in the range 10–15%, a platform approximately 8 m wide should be constructed. The width is reduced to about 5–8 m
for a slope of 15–25%, and about 3 m for a slope of 25–33%. The shoulder bund is constructed with a trapezoidal shape
(1:1 side slope) and a height of 15–30 cm. A bottom width of 75 cm is provided for stability (Figure 25).
Figure 26 shows the typical appearance of bench terraces shortly after construction and some years later.
Worked example (adapted from Mal 1999)
Problem: Design a 120 m long bench terrace for a sandy loam soil with an average slope of 16%. The entire width of
the terrace acts as a channel which is provided with a uniform gradient of 0.6%. Rainfall intensity for a 10 year recurrence
interval and the time of concentration is 24 cm/hr. (The time of concentration is the time needed for water to flow from the
most remote point in the watershed to the watershed outlet.)
Solution: An inward sloping terrace should be used (Figure 25). As the average slope is 16%, the width of the terrace is
selected as 6 m or W = 6 m. A final slope of 5% is provided so that the inner side of the terrace is 30 cm lower than the
outer side.
Average slope (S) = 16%, width of the terrace (W) = 6 m.
The terrace spacing (VI) is given by
VI =
SxW
100 – S
16 x 6 = 1.143 m.
100 –16
The following standard dimensions are assumed: riser side slope = 1:1, shoulder bund height = 30 cm, bottom width = 75
cm, side slope = 2:1
34
The area of the terrace that has to be drained by the channel is calculated using the formula:
A=
LxW
10,000
where
A = area to be drained (ha)
L = length of terrace (m)
W = width of terrace (m)
Thus, the area of terrace, A =
120 x 6
m2 = 0.072 ha .
10,000
Figure 25: Design of (inward sloping) bench terrace (not to scale)
Original land (16% slope)
30
cm
Shoulder bund
1:1 (Riser)
a1
1:1
30 cm
S1
a2
Slope
5.25 m
75 cm
6.0 m
Figure 26: Newly constructed bench terraces on a slope (left) and the same terraces some years later (right)
Source: DWIDP
35
For a sandy loam soil, the runoff coefficient = 0.3. The peak discharge to be handled is obtained from the rational runoff
formula:
Q=
CxIxA
(Mal 1999) ,
36
where
Q = rate of runoff in cubic metres per second,
I = rainfall intensity, that is the rate of rainfall in mm/hr for a designed frequency for a duration equal to the time of
concentration (tc),
A = area of watershed in hectares, and
C = dimensionless runoff coefficient.
Thus,
0.3 x 24 x 0.072
= 0.0144 m3 per second
36
The area of flow can be calculated from Figure 25:
a1 = ½ × 0.30 × 0.30 = 0.045 m²,
a2 = ½ × 5.25 × 0.30 = 0.7875 m².
Q=
The total area of flow is a = a1 + a2 = 0.8325 m².
s1 = (30²+30²) = 42 cm = 0.42 m; s2 (0.30²+5.25²) = 5.259 m,
where s1 and s2 are the perimeter of water flow.
The wetted perimeter (P) is calculated as:
P = s1 + s2 = 5.679 m,
R = a/P = 0.8325/5.679 = 0.147 m.
Using Manning’s formula, the velocity of flow (V) can be calculated from
1
1
2
V = n x R3 x S2
2
1
(0.147)3 x (0.006)2
= 0.538 m/s .
V=
0.04
The velocity is non-erosive. The discharge carrying capacity Q = aV = 0.8325 × 0.538 = 0.45 m³/s. When the terrace
acts as a channel it has sufficient carrying capacity.
Diversions
Diversions are ridges of soil or channels with a
supporting ridge on the lower side. They are built
across the slope to intercept runoff and dispose of
it at a selected location. They are used to break up
long slopes, to direct water away from active erosion
sites, to direct water around agricultural fields or
other sites, and to channel surface runoff to suitable
outlet locations. Safe passage of the surface runoff
to prevent slope failure can be achieved by installing
drainage ditches, or by cross drainage work for road
structures.
Slope drainage
The simplest way to safely drain off springs and
surface water is to use an open ditch (drain) or a
system of open ditches. The main ditch is located
in the direction of the slope gradient (downhill);
secondary or lateral ditches are located in a fishbone
pattern (Figure 27). Water should be collected as
36
Figure 27: Arrangement of ditches on the slope
Tertiary ditch
Main ditch
Secondary ditch
close as possible to its origin and channelled to a side drain, culvert, or any other nearby water course. Ditch
excavation should start at the lowest point and work up in order that the accumulating water may drain off
immediately. The most common types of drain are stone or gravel-filled drains with or without pipes.
Pipe drains are the most efficient and effective, but they are more expensive and often not locally available. Normal
stone drains may silt up over time and it is advisable to form a drainage channel of stones, or place a bundle of
brushwood, at the bottom of the drain. The top of the drain should be covered with a layer of grass to prevent
siltation.
Simple drainage ditches
Drainage ditches can be constructed with different shapes and sizes (Figure 28). The appropriate size and shape for
a particular site depends upon such factors as the expected runoff, site condition, and availability of resources and
construction materials.
Figure 28: Simple designs for drainage channels
Unlined drain
Partially lined drain
Lined drain
Drain made of hollow concrete blocks or bricks
Weepholes
Semi-circular lined drain
Trapezoidal drain with stone masonry lining
Rectangular lined drain
Trapezoidal drain with benching
Source: based on WHO/UNEP 1991
37
Figure 29: Grassed waterway
Filter strips to
trap sediment
Stiff-stemmed plants
in filter strips
Grassed
waterway
Flexible grass
in channel
Source: Modified from Bentrup 2008
Grassed Waterways
Figure 30: Grassed waterway
(adapted from Sharda et al. 2007)
Grassed waterways are natural or artificially
constructed water courses shaped or graded to
the required dimensions and planted with suitable
vegetation (Figures 29 and 30). Grassed waterways
generally run down a slope and are designed to
conduct surplus water safely into natural drainage
courses. They are usually made broad and shallow,
although the shape and size can vary depending on
the size of the drainage area, slope of the land, and
soil type. The channels help surface water to flow
across the land without causing soil erosion. They
are used as outlets to prevent rill and gully formation.
The vegetation in the channel helps control the water
flow and reduces channel surface erosion. Properly
designed grassed waterways can safely transport
large volumes of water to the down slope. They are
also used as filters to prevent sediments entering into
nearby water bodies. Grassed waterways are used as
outlets for diversions and emergency spillways;
Source: Keshar Man Sthapit
to safely convey runoff from contour and graded
bunds and bench terraces;
as outlets for surface and sub-surface drainage
systems on sloping land;
to carry runoff from natural drains and prevent formation of gullies; and
to dispose of water collected in road ditches or discharged through culverts.
The design of grassed waterways is described in Box 8.
Conservation Ponds
Conservation ponds, also known as farm ponds, are small reservoirs constructed for the purpose of collecting and
storing water from surface runoff. Storing water runoff during excessive rainfall helps to reduce the peak flow and
surface erosion and thus reduce the probability of floods. It is also useful for providing supplemental irrigation for
agriculture, water for domestic purposes, and fish farming. Conservation ponds play a significant role in areas with
rainfed agriculture, and construction of a large number of ponds in a catchment area can have a significant effect
38
Box 8: Design of grassed waterways
When designing a grassed waterway, it is first necessary to determine the required shape, size, and gradient.
Shape
Usually a waterway can be triangular, trapezoidal, or parabolic. A parabolic shape is hydrologically more efficient and
easier to construct. The other two shapes tend to become parabolic over time as a result of bank erosion and deposition of
sediments across the channel section.
Size of waterway and calculation of peak discharge
The size of the waterway is calculated on the basis of the peak rate of runoff expected from a ten-year return period storm
without scour or fill. The peak runoff rate can be estimated using the rational method. The rational runoff coefficient varies
according to land use and soil type (Table 4). The catchment area of the grassed waterway increases towards the outlet
and the size of the waterway should be increased accordingly.
The rational equation, in imperial units, is
Q=CxixA
where
Q = peak discharge in cubic feet per second (ft3/s),
C = rational method runoff coefficient,
i = rainfall intensity, inches/hr, and
A = drainage area in acres.
The rational method runoff coefficient (C) is a function of the soil type and drainage basin slope. The rainfall intensity (i)
is typically found from intensity/duration/frequency curves for rainfall events in the geographical region of interest. The
duration is usually equivalent to the time of concentration of the drainage area. The resultant Q can be divided by 35.3 to
convert from ft3/s to m3/s.
Table 4: Rational method runoff coefficients (C) for different land cover areas and soil types
Land use and topography
Soil type
Sandy loam
Clay and silt loam
Tight clay
Cultivated land
Flat
Rolling
Hilly
0.30
0.40
0.52
0.50
0.60
0.72
0.60
0.70
0.82
Pasture land
Flat
Rolling
Hilly
0.10
0.16
0.22
0.30
0.36
0.42
0.40
0.55
0.60
Forest land
Flat
Rolling
Hilly
0.10
0.25
0.30
0.30
0.35
0.50
0.40
0.60
0.60
Populated area
Flat
Rolling
0.40
0.50
0.55
0.65
0.65
0.80
Note: The value of the rational method runoff coefficient can vary from close to 0 to 1.0. A low C value indicates that most of the water
is retained for a time on the ground surface and soaks into the ground, where as a high C value means that most of the rainwater runs
off immediately.
Source: Suresh 1997
Flow velocity
The flow velocity in a grassed waterway depends upon factors such as the soil type, water quality, and ability of the
vegetation to resist erosion. Similarly, the permissible velocity of flow in a waterway varies with the type, condition, and
density of vegetation. The maximum permissible velocities recommended by the Central Water and Power Research Station
(CWPRS) Pune for canals for different types of soils are shown in Table 5. Permissible velocities for sod forming grasses are
higher than those for bunch grasses or other non-uniform grasses. In a grassed waterway, the average flow velocity near
39
the top is always higher than the velocity in contact with
the channel bed, as surface roughness is greater at the
bed.
Table 5: Recommended maximum permissible
velocity for different soils
Soil type
Gradient
The slope of the land normally determines the gradient
of the waterway. Usually, a gradient of less than 5% is
preferred; in a normal course it should not exceed 10%.
Design of the cross-section
The catchment area of the grassed waterway increases
towards the outlet and the size of the waterway should
be increased accordingly. The cross-sectional area is
calculated from the formula
A=Q/V
Maximum permissible
velocity (m/s)
Ordinary soils
0.6–0.9
Very light loose to average
sandy soil
0.3–0.6
Sandy loam, black cotton
soil, and similar soil
0.6–0.9
Hard soil
0.9–1.1
Gravel and disintegrated
rock
1.5
Source: NABARD 2012
where
A = cross-sectional area of flow,
Q = peak discharge, and
V = cross-sectional average velocity.
Q is calculated using the rational method (above) and V is calculated using Manning’s formula to calculate the cross-section
average velocity flow in an open channel:
2
1
V = 1 x R3 x S2
n
where
V = cross-sectional average velocity (m/s),
n = Manning’s roughness coefficient,
R = hydraulic radius (A/P) (m),
A = cross-sectional area of flow (m²),
P = wetted perimeter (m), and
S = slope (m/m).
The value of n can be taken as 0.035, or 0.04 for freshly constructed earthen channels.
Assume either the width or the depth and calculate the dimensions of the channel from A. Check whether the computed
velocity of flow is within the permissible limits. If not, adjust the channel dimensions to bring the velocity of flow within the
permissible limits. Add a suitable freeboard (height between water level and top of the bank) of 20% extra depth or a
minimum of 15 cm to the design depth of the waterway as obtained above to account for any higher flood. on downstream flow and control of floods. Water storage is a topic of increasing importance in the Hindu Kush
Himalayas as the focus turns toward adaptation to climate change, and conservation ponds and other storage
mechanisms are likely to play an increasing role in future development activities (Vaidya 2009; Upadhya 2009).
Conservation ponds can be broadly classified into embankment type ponds and dugout type ponds.
Dugout ponds
The types of dugout ponds range from the very simple, which require no explanation, to forms specifically designed
to collect water on a slope for infiltration and recharge purposes. Some selected types are described below.
Simple earthen ponds are cheap and durable but have high seepage losses that reduce effectiveness. Vertical and
horizontal seepage loss can be significantly reduced by compacting a 30 cm deep layer of heavy clay on the floor
and walls of the pond. Addition of cow-dung and puddling will help seal seepage pores. Watering of buffaloes in
such ponds also helps reduce seepage. High density polythene sheet or SILPAULIN (multi-layered, cross laminated,
ultraviolet stabilized plastic sheet) can also be used to reduce seepage (Figure 31).
40
Figure 31: Earthen ponds laminated with plastic sheet
Source: ICIMOD 2007
Eyebrow pits are a special type of dugout pond used to reduce runoff across a slope and thus stabilize degraded
slopes and increase infiltration and recharge of springs. Small curved trenches around 2 m long and 50 cm wide
in the shape of an eyebrow are dug at intervals facing inward to the slope to catch water and slowly return it to the
soil. Grass and fodder species are planted along the lower ridges of the pits (ICIMOD 2007).
Embankment type ponds
(adapted from Sharda et al. 2007)
Embankment type ponds are constructed to collect runoff at the base of a slope. They can be constructed across
dry water beds or courses with a gentle to steep slope that fill when it rains. Usually, an earthen dam is constructed
between two hillsides to hold back the water from overland runoff. The pond bottom and dam should be made
up of soil that prevents excess seepage. Embankment ponds should not be built by damming any stream with
permanent flow, no matter how small. The dimensions depend on the volume of water to be stored. The location
should be sufficiently depressed to enable the maximum storage volume to be obtained with the least requirement
for earthworks.
The design of embankment type ponds is described in Box 9, and the steps in their construction are elaborated in
Box 10.
Seepage control. Seepage is almost inevitable in any dam or embankment structure. It mainly occurs through the
embankment and under the foundation. The following methods can be used to control seepage.
Core wall: A core wall is constructed using impervious materials and checks the flow of water in the dam. The
main purpose of the wall is to add strength to the bund and increase the seepage gradient length through the
embankment.
Filter and toe drains: These drains are constructed by including a layer of coarse material in the dam section
to attract the flow of water and bring the saturation line down to the level of the drain.
Berm: A berm is a horizontal structure built on the lower part of the downstream (outward) face of the dam to
prevent seepage from the face and keep the seepage line within the base of the dam section. It increases the
base width of the embankment.
Cut off wall: A cut off wall is constructed to join the impervious foundation to the base of the dam and prevent
piping.
Spillway. A spillway is an opening constructed in the embankment to allow water to exit the pond once it is at the
desired level. The size and type of spillway depends upon the size and characteristics of the watershed and the site
conditions. There are two types of spillway: mechanical and emergency. A mechanical spillway is used to let out
41
Box 9: Design of an embankment type pond
Designing the pond involves calculating the dimensions (height, width, side slopes), and considering how to control
seepage and provide a spillway.
Height
The dam height should be no more than 16 m, otherwise special design criteria for stability must be used. The height of the
dam is also called the top bund level (TBL) of the embankment. It is calculated using the formula:
TBL = FTL + flood depth + freeboard + settlement allowance,
where
FTL = the full tank level of the reservoir. This is the required storage volume and is determined by the depth capacity curve
for the location where the dam is to be constructed.
Freeboard = height added to the dam as a safety factor to prevent waves and runoff from storms greater than the design
frequency from overtopping the embankment. Normally, 10–15% of the height is added as freeboard to the highest flood
level of the dam, with a minimum of 50 cm.
The height of the earth fill along the central line of the dam is determined from the cross-section profile of the ground at the
central line and the calculated height of the embankment.
Top width
The top width is calculated from the formula:
W = (H/5) + 1.5,
where
W = the top width, and
H = height of the dam.
In general, a minimum top width of 2.5 m is recommended for dams up to 5 m high.
Side slope
The slope used for the sides depends upon the nature of the fill materials. When fill materials are stable, the side slopes can
be steeper and vice versa. The recommended side slopes for earthen embankments are shown in Table 6.
Table 6: Side slopes for embankments made of different materials
Type of material
Homogeneous well-graded material
Homogeneous coarse silt
Homogeneous silty clay or clay
• Height less than 15 m
• Height more than 15 m
Sand or sand and gravel with clay core
Sand or sand and gravel with reinforced cement concrete (RCC)
core wall
Source: Sharda et al. 2007
42
Upstream slope
Downstream slope
2.5:1
2:1
3:1
2.5:1
2.5:1
3:1
2:1
2.5:1
3:1
2.5:1
2.5:1
2:1
Box 10: Construction of an embankment type pond
Step 1
Completely clear large stones, bushes, and tree stumps from the site of the pond and plough and excavate the top soil to a
depth of 10 cm. Stockpile the excavated soil on the downward (outward) side of the dam. It can be used later to provide a
base for adding grass sod on the downward side.
Step 2
Mark out the layout for the main embankment, core wall, and mechanical and emergency spillways using stakes or lime
powder.
Step 3
Excavate a cut-off trench along the bottom length of the dam as per the design depth, bottom width, and side slopes.
Excavation is generally at least down to the layer of impervious material.
Step 4
Fill the cut-off trench with the best available clay, compacting layer by layer, up to the desired height to form the core wall.
Step 5
Cover the core wall with layers of excavated earth, compacting each layer, to form a bund of the selected dimensions.
Compaction is normally achieved using rollers with pneumatic tyres and sheep foot rollers after slightly moistening the earth
fill to achieve optimum moisture content.
Step 6
Start constructing the mechanical spillway when the embankment construction reaches the bottom level of the spillway
(Figure 32). Take special care to compact the material around the spillway components to avoid unequal settling. Place
pipes, anti-seep collars, and other equipment, and then back fill with earth. Cover the pipe with at least 1 m of earth before
allowing any heavy equipment such as trucks or tippers to pass over it. Proceed in the same way with the emergency
spillway.
Figure 32: Location of mechanical and emergency spillways
Embankment
Crest of emergency
spillway
Crest of mechanical
spillway
Expected flood level
Normal water level
in pond
Gross freeboard
Source: Sharda et al. 2007
43
excess storage water safely; whereas an emergency spillway is used as a safeguard for the earthen embankment
against overtopping when the inflow exceeds the capacity of the mechanical spillway.
Retaining Walls
Retaining walls are artificial structures that hold back soil, rock, or water from a building, structure, or area.
Retaining walls prevent down slope movement and soil erosion, and provide support for vertical or near-vertical
changes in gradient. The walls are generally made from timber, masonry, stone, brick, concrete, vinyl, steel, or a
combination of these. Retaining walls act to support the lateral pressure exerted by a soil mass which may cause
slope failure. Retaining walls are strongly recommended where the toe of slope has collapsed and the slope failure
is likely to progress upward along the slope. Retaining walls should be constructed on a stable foundation. Their
design is described in Box 11.
Retaining walls are categorized in two ways: based on the mechanics of performance, and based on the
construction material.
Types based on the mechanics of performance
Gravity retaining wall. A gravity retaining wall is low and depends on its own weight or mass to hold back the
earth behind it (Figure 33a). It is constructed with a large volume of material in such a way that, when stacked
together, the weight and friction of the interlocking material exceeds the forces of the earth behind. The wall
supports the pressure from the earth by means of its dead weight, and generally requires a good foundation with
sufficient bearing capacity. The wall is thicker at the bottom than at the top; the thickness at the base should be
between one-half and three-quarters of the height. Gravity walls are very cumbersome to construct because they
require large amounts of material. They are usually constructed with concrete and masonry. The size of the section
Figure 33: Types of retaining wall based on the mechanics of performance
Backfill
Backfill
Reinforcement
Backfill
Backfill
Backfill
(a) Gravity retaining wall
(b) Semi gravity retaining wall
Backfill
(c) Cantilever retaining wall
Backfill
Counterfort
Buttress
Backfill
Backfill
(d) Counterfort retaining wall
(e) Buttressed retaining wall
Legend
Dry stone or cement masonry wall
Reinforced cement concrete wall
44
Box 11: Designing a retaining wall
The design of a retaining wall mainly consists of the estimation of the load and active pressure acting on the structure and
the design of the structure to withstand this load and pressure. The forces acting on the wall are the lateral pressure and the
self weight of the wall. The self weight of the wall is responsible for supporting the lateral earth pressure.
Consider as an example the case of a gravity retaining
wall as shown in Figure 34. The wall is supporting the
soil mass with no surcharge.
Figure 34: Gravity retaining wall showing the dimensions
(black arrows) and the forces acting on it (blue arrows)
2
P=
s H (1 sinφ)
2(1+sinφ)
where
P = pressure in kg/m2 (the pressure acts at H/3 above
the base [DSCWM 2005]),
H = height of the wall in metres,
gs = density of the soil in kg/m3, and
f = angle of earth in degrees.
Soil
a
The first step is to estimate the pressure of the earth on
the wall. Rankine’s formula can be used to calculate the
pressure as follows:
C
D
Retaining wall
P
P
X
R
A
F
H
H/3
W
O
E
B
z
b
The weight of the wall per metre length is given by:
W=
(a+b) x H x γm
2
where
W = weight of wall per metre length in kg/m (the weight acts at a distance X from the face BC),
gm = density of wall material in kg/m3,
a = top width of retaining wall in metres, and
b = bottom width of retaining wall metres.
For
•
•
•
the structure to be in equilibrium, the following conditions must be satisfied:
The algebraic sum of all the vertical forces must be zero, i.e., ∑V= 0;
The algebraic sum of all the horizontal forces must be zero, i.e., ∑H = 0; and
The moment of all forces acting on the wall about any point must be zero (to prevent it overturning), i.e., ∑M = 0.
Let RV and RH be the vertical and horizontal reactions at the point of application of the resultant force R on the base of the
wall.
For ∑V= 0, W = RV.
For ∑H = 0, P = RH.
For ∑M= 0, taking the moment about B,
P x H/3+W x X = RH (X + Z),
where
Z = the shift, i.e., EF.
Since RH = W,
P x H/3 + W x X = W(X + Z).
On simplification,
Z= PxH .
3xW
If O is the middle point of base AB, then
eccentricity (e) = OF = BF – BO = X + EF – b/2
= X + Z – b/2.
After calculating the lateral pressure acting on the wall (P), self weight of the wall (W), point of application of lateral
pressure force, centre of gravity of the wall, and eccentricity, the section can be checked for stability.
45
of a gravity retaining wall can be reduced if a small amount of reinforcement is provided near the back face
(Figure 33b).
Cantilever retaining wall. A cantilever retaining wall has a relatively thin stem, usually made of concrete
reinforced with steel to resist the tensile force (Figure 33c). The width of the footing is very important as it is designed
to resist the sliding forces which the earth exerts upon the wall. The wall requires significant steel reinforcing in both
the footer and the wall structures. The steel should extend from within the footer up into the wall so that the two
pieces become one integral unit. This type of wall is generally economical up to a height of 6–8 m.
Counterfort retaining wall. A counterfort retaining wall is similar to a cantilever retaining wall, but further
supported by additional thin triangular shaped walls, or counterforts, built at right angles to the main trend of the
wall. The counterforts are spaced at regular intervals along the wall and connect the back of the wall to the top of
the footing (Figure 33d). The footing, retaining wall, and support walls must be tied to each other with reinforcing
steel. The counterforts reduce the shear force and bending moments in the stem and the base slab and add strength
to the retaining wall. They are hidden within the earthen or gravel backfill of the wall. Counterfort retaining walls are
economical at heights of more than 6–8 m.
Buttressed retaining wall. Buttressed and counterfort retaining walls are similar, with the main difference that in
buttressed walls the vertical brackets are provided in front of the wall (Figure 33e). The buttresses add strength and
help to stabilize the overall wall system. Depending upon the overall length of the main wall, several buttresses can
be constructed at regular intervals.
Types based on construction material
Dry stone masonry. Dry stone masonry walls are usually the cheapest wall structures and are suitable for heights
up to 3–4 m. A skilled mason, suitable stones, and bonding of stones with keystones are required to make a good
quality wall. In general, the width to height ratio varies from 1:1–0.6:1 for walls with heights of 1–4 m.
Gabion wall. A gabion is a heavy duty basket-like structure made in the shape of a box from welded or twisted
galvanized iron wire mesh, divided by wire diaphragms into cells, and filled with heavy material (typically rocks
or broken concrete) that cannot escape through the mesh openings. Gabions are generally used as construction
blocks, and are tied together with galvanized iron binding wire to form larger structures. Gabion walls are
constructed using gabion boxes of various sizes stacked next to and on top of each other before tying. Good quality
stone should be used to fill the boxes, with dimensions preferably not less than 10 cm, or at least greater than the
mesh size. Stones should be packed as tight as possible to increase the density of the gabion wall. The gabion
structures are flexible and provide good drainage due to the dry stone packing.
Cement masonry wall. Cement masonry walls are constructed using good quality stones with cement sand
mortar. These walls are rigid and designed as gravity structures with a base width varying from 0.5–0.75 times the
wall height. The foundation must be on firm, risk-free ground. Weep holes of at least 75 mm diameter should be
included every 2 x 2 m2 in a staggered pattern for drainage. As these walls are rigid and impermeable, they are not
appropriate for construction to hold wet colluvial slopes or where ground movement is expected.
Composite masonry wall. Composite masonry walls are similar to cement masonry walls except that they have
panels of dry stone masonry of about 0.6–1 m square forming a grid on the face and separated by 0.5 m strips of
cement masonry. They are stronger than dry masonry walls but retain the advantage of having relatively good water
drainage.
Cement concrete wall. Cantilever, counterfort, and buttressed retaining walls are constructed with reinforced
cement concrete. The reinforced steel in the wall takes up the tensile stress that the wall is exposed to. The amount
of steel required is calculated by analysing the load on the wall.
Crib wall. A crib wall is a box-type structure built from interlocking struts of timber, precast reinforced concrete,
steel, or other material, and is usually infilled with soil or stone. The whole unit acts as a gravity wall. Due to its
construction without fixed joints, and the segmented nature of the elements, crib walls are flexible and thus to some
extent resist differential settlement and deformation.
46
Safety of a retaining wall
Retaining structures can fail for a variety of reasons. The major types of failure are shown graphically in Figure 35.
The measures used to protect a wall against these different types of failure are summarized in the following sections.
Safety against sliding. Retaining walls should be able to resist the sliding force exerted by the lateral pressure (P)
which tends to cause slide along the plane below the bottom slab, which is resisted by the shear force developed
between the bottom slab and the ground (frictional resistance). If µ is the coefficient of friction between the base
of the wall and the soil, the maximum frictional resistance is equal to µW. Thus for stability against sliding, P must
never exceed µW. The factor of safety (F) is given by
F = µW/P.
A minimum factor of safety of 1.5 is generally recommended.
Safety against overturning. Retaining walls should be able to resist the overturning moment exerted by the
horizontal component of the lateral earth pressure. The resisting moment is composed of the vertical component of
the lateral earth pressure and the self weight.
The section is in equilibrium under the action of four forces (Figure 34):
the horizontal pressure of the earth (P) acting at H/3 from the base of the wall;
the weight of the wall (W) acting at a distance X from the wall face BC;
the vertical component (RV) equal to W and acting at E; and
the horizontal component (RH) equal to P and resulting from the frictional resistance between the wall body and
the ground.
The section can overturn about the point A. As long as the resultant R touches the base, the section cannot overturn.
If R touches the base at A, the section is on the point of overturning, and if it falls outside the base, the section will
Figure 35: Modes of failure of retaining structures
Overturning
Sliding
Bearing capacity failure
Deep seated rotational movement slope failure
Tension failure
47
overturn. Hence, the limiting value is when F coincides with A, i.e., EF = EA, when the balancing moment will have
a value equal to W x EA.
Factor of safety =
Limiting balancing moment
= W x EA .
P x H/3
Overturning moment
The recommended factor of safety against overturning is usually 1.5–2.0.
Safety against bearing capacity failure. The supporting strength of soil or rock is referred to as its bearing
capacity. The maximum pressure which soil can carry safely without risk of shear failure is the safe bearing capacity.
In order to avoid bearing capacity failure of the soil at the base, the maximum comprehensive stress acting normal
to the base must be less than the allowable bearing capacity of the soil. The maximum comprehensive stress
normal to the base must also be less than the maximum comprehensive stress for the masonry to avoid crushing the
masonry at the base.
Factor of safety =
Allowable bearing pressure or permissible comprehensive stress for the masonry .
Maximum comprehensive stress at the base of wall
A factor of safety of 3 is recommended for safety against bearing capacity failure.
Safety against tension failure. There should be no tension at the base of the wall. To avoid tension within the
structure, the eccentricity (e) should be not more than b/6 on either side of middle of the base, i.e., at the point O.
Under such conditions, the resultant, R must be within the middle third of the base width.
In designing a gravity type retaining wall, a trial section is chosen and checked for all the stability conditions
mentioned above. If the stability checks yield unsatisfactory results, the section is changed and rechecked. Table 7
shows the design dimensions for a typical retaining wall structure.
Table 7: Typical retaining wall design specifications
Type
Dry stone
masonry
Composite
masonry
Cement masonry
Top width
0.6–1.0 m
0.6–1.0 m
Base width
0.5–0.7 H
Front batter
Back batter
Gabion
Low
High
0.5–1.0 m
1.0 m
1–2 m
0.6–0.65 H
0.5–0.65 H
0.6–0.75 H
0.55–0.65 H
vertical
varies
10:1
6:1
6:1
varies
vertical
varies
varies
varies
Inward dip of
foundation
1:3
1:3
horizontal or 1:6
1:6
1:6
Foundation depth
below drain
0.5 m
0.5–1 m
0.5–1 m
0.5 m
1m
Height range (H)
1–6 m
6–8 m
1–10 m
1–6 m
6–10 m
Hill slope
<35°
20°
35–60°
35–60°
35–60°
Cement 50 cm
thick, masonry
bands at 3 m
centre to
centre
Make weep holes of
75 mm diameter at
1–2 m c/c. Provide 50
cm rubble backing for
drainage.
Hand pack stones. Select block
shapes in preference to flat. Specify
maximum/minimum stone size. Do
not use weathered stone. Compact
granular backfill in a layer (<15 cm).
Toe protection in case Boulder pitching
of soft rock/soil
General
Set stones along
foundation bed.
Use long bond
stones.
Foundation to be stepped up if rock encountered
All walls require durable rock filling of small to medium size.
Drainage of wall bases not shown
Application
Least durable
Most durable
Non-ductile structures, susceptible to earthquake damage
48
Can adjust to settlement and slope
movement
Very flexible structures
Drop Structures
Drop structures, also known as grade control structures, are structures placed at intervals along a channel reach
to change a continuous steep slope into a series of gentle slopes and vertical (or steep and roughened) drops, like
a series of steps. They control erosion and river channel degradation by reducing the slope of the channel and
preventing the development of high erosive flow velocities, and allow water to drop safely from one level to another
without gouging out gullies. They can also help to control flooding and trap the sediment moving with runoff water.
Drop structures include sills, weirs, chute spillways, drop pipes, and check dams. A weir allows water to run over the
edge like a miniature waterfall, dropping down onto a concrete apron. The apron absorbs the impact of the falling
water and then the water streams to an outlet. When the drop in grade is more dramatic, a chute can be used to
prevent severe erosion. As the name implies, water moves down a chute made of concrete or lined with rocks or
concrete blocks. Like chutes, pipes are effective in handling water when the drop in grade is dramatic. They are
designed to carry water through or under an earth embankment to a lower elevation. With a drop inlet, water drops
down into the inlet and then flows through the pipe. Because of the high energies that must be dissipated, preformed scour holes or plunge pools may be required below these structures.
In steep hill and mountain areas, the most common drop structures are check dams, which are used to control gully
erosion. Check dams are described in more detail in the following section. Drop structures are also used to reduce
the effective slope below the upper limit for a grassed waterway. Weirs and sills are more common downstream as
river training measures.
Drop structures can be made of concrete, timber, sloping
riprap sills, and soil-cement or gabions. Drop structures made
from timber or logs are more appropriate in small streams and
gullies.
Figure 36: Gully erosion (top) and check
dam treatment (bottom) in Dahachowk,
Kathmandu, Nepal
EC (1997) contains more details on the implementation of some
of these structures.
Check dams
Check dams are small low drop structures built across a gully or
channel to prevent it from deepening further. These small dams
decrease the slope gradient and reduce the velocity of water
flow and the erosive power of the runoff. They also promote the
deposition of eroded materials to further stabilize the gully or
channel.
Gully plugging using check dams, accompanied by planting
between the dams to stabilize the channel, can be one of the
most effective ways to conserve soil and water and rehabilitate
land degraded by gullies (Guedel 2008). The effectiveness of
different check dams depends upon the design, location, and
construction materials. Figure 36 shows an example of check
dams constructed to reduce gully erosion.
Check dams can be constructed from a wide range of
materials including rock, wood, bamboo, gravel bags, sand
bags, concrete, masonry, and fibre rolls. The characteristics,
advantages, and disadvantages of some major different types of
check dams are summarized in Table 8. Two different types are
shown in Figure 37.
Details on check dam design are given in Box 12, and
considerations for their construction are given in Box 13.
Source: DWIDP
49
Figure 37: Gabion check dam (left) and masonry check dam (right)
Source: Peng Huang
Source: DWIDP
Table 8: Characteristics of some common types of check dam
Type of
check dam
General characteristics
Brushwood Made of wooden poles and brush
Least permanent of all types if not
rooted
Uses local materials
Low cost
Commonly used in gully control where
boulders or rocks are abundant
Takes a long time for the dams
to develop roots and become
If roots and shoots develop, they
established
can form a long-term barrier
Simple
Low cost (where stones are
abundantly available)
If not made properly or stones are
too small, they can be washed
away
Made of big boulders or rocks
Stability and strength depends on
the size of the boulders or rocks and
quality of construction
Simple
If properly made, are almost
Commonly used in gully control where permanent and durable
boulders or rocks are abundant
If large voids are not properly
filled they, may create water jets,
which can be destructive if directed
towards the bank
Made with wire gabions of different
sizes filled with stones
Flexible and permeable
More expensive than loose stone or
boulder structures
Low cost
Transportation of big boulders is
difficult (if not available upslope of
the site)
Flexible
Suitable where the land mass is
unstable
Preferred where big boulders are not
available
Economical compared to other
solid structures
The gabions have to be brought
from outside which increases the
cost
Masonry
Made of cement masonry or concrete
Permanent solid structure
High cost
Generally only used to protect
important infrastructure such as roads
and buildings
Good appearance
Materials not locally available
(cement, rods)
Source: DSCWM 2004
50
Simple
Low cost where materials are locally
available
Stability and strength depends on
the size of rocks and quality of the
construction
Gabion
Disadvantages
Suitable for small gullies 1–2 m deep
Loose stone Made of loose stone or rock
Boulder
Advantages
Need skilled labour for construction
Need more engineering design,
and skilled labour for construction
Box 12: Designing a check dam
The discharge rate through the channel must be calculated first. The hydraulic element of the design, especially the spillway
section, is also very important as any fault in the hydraulic design can reduce the life of the structure.
Spacing
The spacing between dams is an important factor in the design (Figure 38). The space between consecutive check dams is
selected to obtain the desired gradient between the bottom of the upper dam and the top of the lower dam – known as the
compensation gradient. The spacing depends on the slope of the original waterway, the compensation gradient, and the
effective height of the dams (DSCWM 2005). It is given by
d=hx
100
S0 Se
where
d = spacing between two successive check dams (horizontal distance),
h = height of the check dams up to the notch,
S0 = existing slope of bed in per cent, and
Se = stabilizing slope of bed in per cent (usually 3–5%).
The number of check dams (N) is calculated as follows (DSCWM 2005):
N=
a
H
b
where
a = total vertical distance between the first and the last check dam in that portion of the gully or torrent,
b = total vertical distance calculated according to the compensation gradient for that portion of the gully, and
H = average height of the dams.
Runoff estimation
Various methods are used to estimate the runoff rate. The rational formula is the simplest method for determining peak
discharge from drainage basin runoff, but the calculation is only possible if the rainfall intensity, area of watershed, and
runoff coefficient are known (DSCWM 2004).
Q=
C x Itc x A
360
where
Q = rate of runoff in m3/s,
Itc = rainfall intensity in mm/hr for a designed frequency and a duration equal to the time of concentration (tc),
A = area of watershed in ha, and
C = dimensionless runoff coefficient.
Figure 38: Check dam spacing
Se
h
Compacted earth
d
Second check dam
h
S0
Existing slope of gully bed %
Compacted earth
First check dam
51
The time of concentration (gathering time) is calculated from
tc =
L 1.15
15 x H 0.38
where
tc = time of concentration (gathering time) in hours,
L = length of the watershed along the main stream from the outlet to the most distant ridge in km, and
H = difference in elevation between the watershed outlet and the most distant ridge in km.
Note: The rational formula can only be used when a rainfall intensity (Itc) map of the given area is available with
frequencies of 5, 10, 25, 50, and 100 years. If no is map available, the following discharge formulae must be used
instead (Kresnik’s run-off equation and Manning’s velocity or runoff rate formula).
The Kresnik run-off equation is
Q max = C x A½,
where
Q max = maximum permissible discharge in m3/s,
A = catchment area of the gully above the proposed check dam in km2, and
C = coefficient ranging from 0.6–2.0 (depending on land use type).
Note: the Kresnik equation gives the best results for gullies with catchment areas of less than 20 ha. It can also be used in
torrent control for catchments up to 300 ha.
The Manning formula estimates the runoff rate from the river bed characteristics:
2
1
1 x R3 x S 2
V= n
where
V = velocity of flowing water (m/s),
n = roughness coefficient of the channel (for gully channels, n can be set at 0.025),
S = gradient of the gulley channel (%),
R = hydraulic radius (wetted area divided by wetted perimeter) (m) or R = A/P,
A = cross-sectional area of the river (m2), and
P = wet surface of the river (m).
Note, however, that this formula is not accurate for rivers with a high bed load or with mudflow, because this changes the
specific weight of the water.
The notch of the check dam, i.e., the spillway section, is designed to allow the spillway to accommodate peak runoff. The
dimensions are calculated from
Q=C×L×D
2/
3,
where
Q = maximum discharge of the gully catchment at the proposed check dam point (m3/s),
C = coefficient, 3.0 for loose rock, boulder, log, and brushwood dams; 1.8 for gabion and cement masonry dams,
L = length of spillway (m), and
D = depth of spillway (m).
Foundation depth
The check dams are built on a foundation which anchors them into the ground to increase stability and ensure that they do
not collapse or overturn when the peak flow or run off occurs or the dams are silted up. The following should be taken into
account in the design and construction of the foundation:
• The bottom of the foundation should lie below the scour level.
• In erodible strata, if D is the anticipated maximum depth of scour below the designed highest flood level, including
possible concentration of flow, the minimum depth of foundation below the highest flood level should be 1.33D.
• The scour depth should be taken from the expected bed level after siltation of the lower check dam and establishment
of the new bed gradient, due to the reduced bed load after the erosion control.
• As a rule of thumb, take the foundation to be 1 m.
Scour depth
The safety of the check dams is mostly endangered by scouring. Scour occurs when the bed velocity of the stream reaches
the velocity that can move the particles of the bed material. The scouring action of the current is not uniform; it is deeper at
the obstruction and at bends.
52
The scour depth is calculated from Schocklitch’s formula (DSCWM 2004):
Scour depth (Ds) = (4.75 × h0.2 × q0.57)/dm0.35
where
Ds = scour depth in m below water level,
dm = grain diameter in mm, determined on the basis that 90% of the bed material is smaller than dm,
h = water level difference in m above and below the check dam, and
q = runoff in m3/m width of spillway.
The breadth of the scour hole is calculated as 1.5 × length of the notch.
2/ 1.5
3)
The length of the scour hole or apron is calculated as 4 × (0.467 × q
× h0.5.
General design considerations
Check dams are designed for safety against overturning, safety against sliding, and safety against the bearing pressure on
the foundation soil. Table 9 summarizes the suggested general design specifications for different types of check dam.
Table 9: General specifications for check dams
Dam type
Maximum effective
height
Minimum
foundation depth
Thickness
of dam at
spillway level
Slope of the
Slope of the
downstream
upstream face
face of the dam of the dam
Thickness of
the base of the
dam
Brushwood
1 m from ground
level
0.75–1 m
–
–
–
–
Loose stone
1.0 m
0.5 m
0.5–0.7 m
20% (1:1/5)
vertical
calculated
accordingly
Boulder
2.0 m
half of effective
height
preferably
1.0
30% (1:3)
vertical
calculated
accordingly
Gabion
may vary
(recommended not
more than 5.0 m)
half of effective
height
>1 m
20% sloped,
stepped, or
vertical
stepped or
vertical
calculated
accordingly
Note: Use of these dimensions means no stability test is needed against overturning, collapsing, or sliding. However, the size of the
spillway needs to be calculated according to the maximum discharge of the gully watershed area.
Source: DSCWM 2004 (based on Geyik 1986)
Box 13: Basic considerations for check dam construction
The function of check dams is to reduce the gradient and minimize the hydraulic energy of the flowing water. The flow
velocity, and thus the erosion capacity, is controlled by the size of the dams and spillways. The following should be
considered when constructing check dams.
• Construction should normally start at the downstream end of the active section of a gully.
• The top of the check dam should be below the level of the adjacent land to prevent spillover of flood to either side of
the gully.
• The height of the outflow should be no more than 1 m. The lower the check dam, the smaller the risk of collapse and
overflow to the side and of the need for repair. The ideal height of the spillway is often 0.5–0.6 m. It is better to build
two low dams in a cascade than one high dam.
• The check dam should be made lower at the centre to form a spillway. The spillway will draw the stream to the middle,
thus hindering erosion of the gully sides.
• The check dam should extend into the gully floor and the sides of the gully. This ‘keying in’ will help prevent erosion
scouring and tunnelling under or around the check dam. The depth of the keying in depends on the local soil
conditions, but should usually be between 0.3 and 1.0 m, more when the sides of the gully are unstable.
• An apron should be constructed in the gully area immediately downstream of the check dam to protect it from the
erosive forces of the falling water. Without an apron, the check dam will be undercut and will eventually collapse. The
apron should be 1.5–2 times longer than the height of the check dam depending on the slope of the gully. The greater
the slope, the greater should be the length of the apron. Both the floor and the sides, from the top of the spillway to the
downstream end of the apron, should be protected by piling stones.
53
Sabo Dams
Figure 39: Debris flow controlled by a sabo dam
(adapted from Ikeda 2004)
Sabo dams are a common measure to limit debris
flows. They are in some ways similar to check dams,
but they are intended to limit debris flow rather than
runoff velocity. The word sabo comes from Japanese
and means soil conservation (sa means soil and bo
means conservation); sabo dams are a Japanese
technology that is now becoming popular beyond
Japan. Sabo dams are relatively small structures built
across the river bed in upstream areas in the form of a
cross dike. They look like a normal small dam, except
that they have a lower ‘open’ section at the centre
which allows debris to pass through during normal
Source: Sundar Kumar Rai
conditions but prevents large-scale debris flow during
flash floods. Sabo dams built in the upstream areas of
mountain streams accumulate sediment and suppress
the production and flow of sediment. Those built at
the exits of valleys work as a direct barrier to a debris flow which has occurred. A sabo dam with slits is particularly
effective in capturing a debris flow because it has a larger capacity of sand pool under normal conditions.
Sabo dams are usually constructed using masonry, concrete, reinforced concrete, or steel cribs according to the
conditions in the planned area (Figure 39). The main functions of a sabo dam are to
reduce erosion of the river bed and bank;
trap sediment discharge;
control sediment discharge;
grade the effect of sediment; and
reduce the energy of debris flows.
Based on their purpose and the way they function, sabo dams are classified into four types:
check consolidation dam;
river bed erosion control dam;
river bed sediment runoff control dam; and
debris flow control dam.
Sabo dam design is described in Box 14.
Gully Control
Gullies are a highly visible form of soil erosion created by running water. They are deep-sided water courses,
metres to tens of metres in depth and width, gouged out by surface water flow. Gully formation can start with the
formation of rills by surface water flowing down a slope, especially where soil is exposed (Figure 41). Water flow is
concentrated and accelerated down the rills, leading to increased erosion and eventually formation of a full gully.
Gullies channel and accelerate runoff, and thus contribute to flood and flash flood development, as well as causing
damage to the surrounding area and infrastructure, reducing the productivity of farmland, and contributing to
sediment flow and sedimentation of downstream lands, streams, channels, and reservoirs (DSCWM 2004).
The major causes of gully erosion are:
erosion in the catchment,
channel erosion by unmanaged runoff resulting in downward or sidewards scoring,
steep unprotected slopes and drainage channels,
gully head expansion, and
side slope failure due to toe cutting of the gully channel embankment (DSCWM 2004).
54
Box 14: Designing a sabo dam
Sabo structures must be designed according to the intended function and purpose and should be stable enough to
withstand all the expected design forces. The main steps in designing a sabo dam are as follows:
• determination of general design considerations,
• determination of design parameters of debris flow,
• design of open section,
• design of dam body,
• design of dam foundation, and
• design of other appurtenances.
The basic principles of design are summarized in the following. More details can be found in Ikeda (2004).
General considerations
The location and height of the dam are chosen according to the dam’s purpose. The height may also be restricted by the
geological and topographical conditions at the proposed location. The dam should be located on stable ground, if possible
on firm bedrock, as dams are easily destroyed by slope failure if constructed in an unstable location. Ideally the dam should
be located in a stable narrow section of the riverbed to give the highest stability and most cost effective construction. The
height of the dam is usually reduced if it has to be built on a gravel base.
Design parameters of the debris flow
Before designing the dam, it is important to determine the design parameters of the debris flow. The fluid dynamic force of
the debris flow, peak debris flow discharge, velocity of the debris flow and maximum water stage, height of the debris flow,
and density of the debris flow should all be calculated or estimated.
Designing the open section
The topography and geological features upstream and downstream of the sabo dam should be taken into consideration
when designing the open section. The axis of the sabo dam is placed at right angles to the direction of the river with
the open section at the centre of the river course. The open section should be at least 3 m wide, and the final width also
depends on the width of the stream bed. The height of the open section is determined from the design depth of the opening,
the freeboard, and the maximum diameter of boulders expected in the debris flow.
A 50% sediment discharge is added to the actual flood discharge to obtain the design discharge.
Q = (1+0.5) Q1,
where
Q = design discharge, and
Q1 = actual flood discharge.
Design of the dam body and foundation
The design of the body of the dam is based mainly on a stability analysis. The main forces acting on the dam body are
overturning, sliding, and bearing resistance of the foundation. The debris flow hydraulic forces are calculated by assuming
a unit weight of debris-laden water of 11.8 kN/m3.
The dam foundation is determined by considering the bearing capacity and the nature of the underlying foundation material
(soil or rock). Foundation treatment such as construction of a cut-off wall or slurry wall is recommended when the material is
poor. A cut-off wall should be constructed at the toe of the dam to prevent damage by scouring.
The
•
•
•
•
•
design loads that must be considered for a gravity type sabo dam are:
hydraulic static pressure,
sedimentation pressure,
uplift pressure,
seismic inertia force, and
hydrodynamic pressure during an earthquake.
Sub-sabo dam
A sub-sabo dam can be constructed when the main dam height is high or the overflow depth is deep (Figure 40). The main
function of a sub-sabo dam is to reduce the overflow energy of the mass overflowing the dam. Selection of the position of
the sabo dam and sub-sabo dam should be based on the extent of sedimentation and designed to prevent sediment-related
disaster.
55
The desirable distance between the main dam and sub dam can be calculated using the following empirical formula.
L = β x (H1 + h3),
where
L = distance between the main dam and sub dam,
β = 1.5~2.0 (value depends on height of sabo dam at overflow section and width of crest),
H1 = height of the main dam (front apron) above the bedrock (m), and
h3 = overflow depth of the main dam (m).
Figure 40: Main sabo dam and sub-sabo dam
h3
Direction of flow
Main dam
H1
Sub-dam
Face of apron
Face of base rock
L
Figure 41: Gully formation
Source: ICIMOD 2008
56
Gully expansion mainly results from headward erosion by concentrated surface water flow. As a result, the gullies
often expand upslope.
Ideally, land protection measures should be implemented across a catchment to help prevent gully formation. Once
a gully is formed, however, remediation treatment should be undertaken to prevent further expansion of the gully
and to slow the flow of runoff downstream. The main approaches are summarized as follows.
Improvement of the catchment above the gully
The techniques for improvement of the catchment above a gully are the same as those recommended to prevent
gully formation. Essentially, they include all the methods recommended for controlling surface runoff by increasing
infiltration and trapping moisture, as well as soil protection measures and slope stabilization. Measures include
bioengineering approaches such as contour farming, strip farming, mulching, and afforestation, as well as physical
measures such as bunding, terracing, levelling, and trenching, and building structures for diversion of surface runoff
such as grassed waterways and weirs. Details of some of these are given in earlier sections.
Stabilization of the gully head
It is important to stabilize the gully head and prevent it from advancing. This can be done by diverting the runoff to
stop it from entering the gully, or by allowing the runoff from upstream to enter the gully safely by installing drop
structures.
Diversion of surface water above the gully head
In some cases, diverting excess runoff away from the gully head can control gully erosion. For this, diversion drains
are constructed above the gully head to divert the excess runoff away from the gully.
Check dams
The major approach used to treat a gully is to build check dams to reduce the flow velocity down the gully. This
method is described in detail in earlier sections.
Integrated measures for gully stabilization and treatment
Gullies are treated using a range of the bioengineering and construction measures outlined above in the watershed
above the gully head, around the gully head, and down the gully. The aim is to prevent further deepening and
widening of the gully and to control the flow velocity. Table 10 summarizes the different remediation techniques
recommended for different depth classes of gully. Table 11 summarizes the criteria for the selection of different
control measures along different parts of a gully (Geyik 1986).
Table 10: Remediation techniques for different sizes of gully
Type
Depth
(m)
Recommended remediation techniques
Waterway
Gully area
Rill
<0.30
Fascines (community land), conservation
tillage (agricultural land)
Planting
Small gully
0.3–1
Palisades, brushwood, check dams
Planting/brush layering/wattling
Medium gully
1–5
Brushwood and loose stone check dams
Planting/brush layering/wattling
Large gully
5–10
Loose stone and gabion check dams
Planting/brush layering/wattling/retaining walls
Ravine
>10
Loose stone and gabion check dams
Planting/brush layering/wattling/retaining walls
Source: DSCWM 2004
57
Table 11: Control measures for different parts of a gully
Length of main
gully channel
Gradient of main
gully channel (%)
Catchment
area of gully
Required structural measures for each portion of main gully
channel
–
> 2 ha
Above gully head: Diversion ditches or channels
Various
> 2 ha
Up to 100 m from gully head: Brush fills, earth plugs, wovenwire, brushwood, log, and loose stone check dams.
70 or less
2–20 ha
Between 100 and 1,000 m: Boulder check dams, retaining
walls between check dams if necessary; first check dam is
usually gabion or cement-masonry
–
100 m or less
(from gully head)
> 900 m
Note: All structural measures should be accompanied by vegetative measures (planting of tree seedlings and shrub and grass cuttings;
sowing of tree, shrub, and grass seeds).
Source: Geyik 1986
Trail Improvement
(adapted from DSCWM 2004)
Notwithstanding the advantages of road and trail construction in hill areas, it also poses considerable problems.
The combination of slope instability, lack of understanding of slope dynamics, and poor planning and construction,
means that roads and trails are a major source of landsides, slips, and flows in many parts of the Himalayan region,
and thus contribute to the development of flash floods. Operations such as blasting and chipping create geological
disturbances in the rock and soil masses of the mountain slopes. Blasting operations exert a tremendous dynamic
force causing the movement of slip zones, cracks, fissures, and weak planes. In addition, poor stabilization of bared
slopes and trail edges combined with heavy rainfall may lead directly to slips of different types originating along the
line of the trail itself.
Trail improvement refers to the vegetative and structural measures used to protect trails from erosion and to improve
them for people and livestock traffic, both during construction and in the form of remedial measures. General
guidelines should be followed to ensure slope safety when designing and constructing trails and roads along steep
slopes. Detailed discussion of the construction of roads and trails is beyond the scope of this manual but there are
many publications available that the engineer can refer to, including the Mountain Risk Engineering Handbook
published by ICIMOD in 1991 (ICIMOD 1991a,b) and various publications by the Swiss and German development
agencies on ‘green road’ construction (GTZ and SDC 1999; SDC 2008).
Box 15 provides some basic guidelines for ensuring that trails (not roads) avoid negative impacts on slopes and
runoff, together with some suggestions for remediation of specific problems.
Existing trails may start to show signs of erosion damage and instability which need to be addressed through
remediation measures. Table 13 summarizes some of the improvements recommended for slopes of different
steepness and condition.
Table 13: Trail improvement on different slope
Slope
Condition/problem
Suggested improvements
Flat
Earthen, no problem seen
Protection of the surrounding vegetation
<8°
Earthen, with rills
Rill plugging with fascines, palisades, sowing grass, construction of cross drains,
levelling of the trail
<12°
Earthen, with gullying
Stone paving, construction of cross drains, planting at water disposal sites
<12°
Trail turns into a waterway
Stone paving with a concave profile and drop structures (e.g., checkdams,
stepped falls), protection at water disposal sites
>12°
Earthen, with gullying
Stone steps, sowing grass, drainage management
Source: DSCWM 2004
58
Box 15: Responsible trail design
Basic design considerations
Table 12: Recommended step size for different
• Ideally, trails should follow a contour.
slopes
• All conservation measures required for the trail and
slide slopes should be designed and implemented as
Slope (θ)
Riser height (R)
Tread length (T)
a package.
(degrees)
(m)
(m)
• Drainage ditches should be provided at appropriate
12–16
0.12
0.35
locations to guide surface runoff.
17–20
0.15
0.35
• The trail should slope outwards. A maximum cross
slope of 1:20 (vertical height to horizontal length) is
21–23
0.15
0.30
recommended to avoid cross ruts.
24–26
0.17
0.30
• Trails should be wider than 1.2 m.
27–30
0.20
0.30
• An average gradient of 10% is generally considered
>30
0.20
0.25
to be the maximum for comfortable walking; 15% is
considered to be the maximum permissible gradient.
Source: DSCWM 2004
• Trails with gradients of less than 8º (≈14%) should
be cut and levelled and sown with grass.
• Trails with gradients of 8º to 12º (≈20%) should be paved with stone.
• Stone steps should be constructed on trails with gradients above 12º (≈20%). The recommended step size is given in
Table 12.
• The length of the landing (step) can be 1 m.
DSCWM (2004) gives detailed instructions for designing steps.
Practical recommendations
•
•
•
•
•
•
•
Safe disposal of trail runoff is one of the keys to the control of erosion associated with the trail.
Trail runoff can be stored in conservation ponds for later use.
Line the trail with stones wherever possible to reduce erosion.
Trails can be used as waterway channels during the monsoon, make the stone pavement concave, and ensure the
speed of the runoff does not exceed the resistive force of the lining material.
On sloping sections where water may run down the line of the trail, install open, stone-lined cross drains at intervals
of 10–15 m. These should lead off to a soak away area, into a well-vegetated line of natural drainage, or into a
collection pond.
Plant double lines of thorny shrubs along the side of the footpath to discourage people and animals from straying off
the trail and eroding the sides.
Plant trees, shrubs, or grass in nearby areas, improve drainage, and treat rills and gullies wherever necessary.
59
Chapter 6: Physical Methods
for River Training
Flash flood mitigation in the upstream part of a catchment is aimed at reducing the occurrence of flash floods
and focuses on reducing slope instability, reducing the amount and velocity of runoff, and preventing erosion. In
the downstream areas, the focus is on mitigating the effects and impact of any flash flood that occurs. Some rivers
are particularly prone to flash floods (‘flashy rivers’) and it is possible to plan mitigation interventions, even though
the timing of individual flash floods cannot be predicted. This chapter looks at some structural measures in the
downstream areas.
The morphology of a river is a strong determinant of flow, and can thus serve to intensify or mitigate flood waves
and torrents. At the same time, when rivers flow in an alluvial plain they often become meandered or braided, and
at times of flood, this morphology leads to excessive bank cutting which can destroy agricultural land and human
settlements.
‘River training’ refers to the structural measures which are taken to improve a river and its banks. River training
is an important component in the prevention and mitigation of flash floods and general flood control, as well as
in other activities such as ensuring safe passage of a flood under a bridge. For flash flood mitigation, the main
aim is to control the water discharge regime in the watercourse by limiting its dynamic energy, thereby controlling
the morphological evolution of the watercourse (Colombo et al. 2002). River training measures also reduce
sediment transportation and thus minimize bed and bank erosion. Many river training structures are implemented
in combination with bioengineering techniques to lessen the negative effects on environment and landscape (see
Chapter 3). There are a number of types of river training structure. The selection and design of the most appropriate
structure depends largely on the site conditions.
River training structures can be classified into two main categories: transversal protection structures and longitudinal
protection structures.
Transversal Protection Structures
Transversal protection structures are installed perpendicular to the water course. They are used to lower the river
gradient in order to reduce the water velocity and protect the river bed and banks from erosion. Most of the rivers in
the Hindu Kush Himalayan region originate in the high mountains, where they have steep gradients giving the flow a
massive erosive power. Moreover, intense rainfall and breakout events can accelerate the river flow to such an extent
that the water has a significant impact on the watercourses and surrounding areas. Transversal protection structures
are effective for controlling the velocity of rivers and streams and reducing the development of flash floods. The
major structures likely to be useful in the region are described briefly in the following.
Check dams
Check dams are described in detail in the previous chapter, mainly in relation to gully control. The dams used
along river courses follow the same principles. They can be made of gabions, concrete, logs, bamboo, and many
other materials. These dams decrease the morphological gradient of the torrent bed and reduce the water velocity
during a flood event by increasing the time of concentration of the hydrographic basins and reducing the flood peak
and solid transportation capacity of the river. They also help to reduce erosion and debris flow. The main purpose
of check dams on rivers is to stabilize the riverbed over a long distance. Check dams generally require additional
protection structures in the bed or on the banks to hinder undermining.
60
Spurs
Figure 42: Gabion spurs with sandbags along the bank
A spur, spur dyke, or groyne is a structure made to
project flow from a river bank into a stream or river
with the aim of deflecting the flow away from the side
of the river on which the groyne is built. Two to five
structures are typically placed in series along straight
or convex bank lines where the flow lines are roughly
parallel to the bank (McCullah and Gray 2005).
Spurs help train a river to flow along a desired course
by preventing erosion of the bank and encouraging
flow along a channel with a more desirable width and
alignment (Julien 2002). They are used to control
natural meandering at a river bend, to channel wide
rivers, and to convert poorly defined streams into well
defined channels. The spurs create a zone of slack
flow which encourages silting up in the region of the
spur to create a natural bank. They generally protect
the riparian environment and often improve the pool
habitat and physical diversity.
Source: Mercy Corps
Figure 43: Low-cost bamboo and stone spurs
Spurs can be made from many materials including
stone, for example in the form of gabions (Figure 42)
or in bamboo ‘cages’ (Figure 43); tree trunks and
branches (Figure 44); concrete; or any material that is
not easily detached by the river and is strong enough
to withstand the flow and the impacts of debris. They
can be categorized on the basis of permeability
(Figure 45), submergence (Figure 46), orientation
(Figure 47), and the shape of the head (Figure 48).
Some guidelines for designing and constructing spurs
are provided in Box 16.
Sills
A sill (also called a bed sill or ground sill) is a
transverse gradient control structure built across the
bed of a river or stream to reduce bed or headward
erosion. Sills are installed along river stretches with a
medium to low morphological gradient. The purpose
is similar to that of a check dam, but a sill is much
lower. A sill is usually constructed together with other
hydraulic structures such as bridges to prevent them
from being undermined and increase their durability.
Sills can be built with different shapes, for example
stepped or sloping, and from a variety of materials
including concrete, stone, gabions, wood, and rock.
The selection of material depends on morphological
and ecological factors. Sills made from wood, rock,
and gabions tend to be more environmentally friendly
than those made from concrete or cemented stones.
The most common types of sills are the following.
Source: Deepak, DSCWM
Figure 44: Low-cost spurs made from tree trunks
and branches
Source: Mercy Corps
61
Figure 45: Permeable and impermeable spurs
Typical flow
Permeable
Impermeable
Figure 46: Submerged and non-submerged spurs
Water surface
Submerged
Non-submerged
Figure 47: Different orientation of spurs
Typical flow
Attracting
Deflecting
Repelling
Figure 48: Different shapes of spur
River flow
River flow
River flow
River flow
River flow
River bank
River bank
River bank
River bank
River bank
Straight
Mole head
L-head
T-head
Hocky stick shaped
Concrete or stone sills. Sills made of concrete or concreted stone are easy to construct and relatively common,
even though the construction cost is generally higher than for other types. This type of sill can be used for a
wide range of morphological conditions, and is particularly suitable for lower reaches. They are often used in
combination with structures such as bridges or walls.
Gabion sills. Sills made with gabions can be installed under many different hydrodynamic conditions. The gabions
can be filled with rock from along the river or stream bed. Gabion sills are considered environmentally less harmful
than concrete sills for the natural riverine environment and ecology because of their greater width and limited
height.
Wood and rock sills. Sills are often made of local wood and rock in the mountainous reaches of watercourses
or at sites with morphological constraints. Any kind of water resistant wood can be used, the most suitable being
chestnut, larch, and natural or treated resinous plants. This type of sill has a low environmental impact because of
its tendency towards naturalization, which favours the ecology and environment of the watercourse.
62
Box 16: General guidelines for spur design and construction
Spurs should not be used where the river is already narrow or where the alignment of the river banks cannot be modified
or reduced. It is also not advisable to use spurs where the opposite bank is exposed to transverse flows, which create
unacceptable erosion. In such cases continuous longitudinal protection is required (Maccaferri Australia n.d.).
The effectiveness of a spur depends on its design and location, and the resources available. The location of the upstream
starting point and the downstream termination point also influence the success of spur installation. The main characteristics
to be considered are summarized in the following.
Permeability
Spurs can be permeable or impermeable (Figure 45). Impermeable spurs are built of local soil, stones, gravel, rocks, and
gabions, while permeable spurs usually consist of one or several rows of timber, bamboo, or similar. An impermeable spur
blocks and deflects the river flow, while a permeable spur allows water to pass through but reduces the water velocity.
Spur height
Spurs can be designed to be higher than the water level at all times (non-submerged), or submerged during the time of
floods, emerging only when the flood recedes (Figure 46). In general, submerged spurs are designed to be permeable,
whereas non-submerged spurs are impermeable.
The height of non-submerged spurs should not exceed the bank height because erosion at the end of the spur in the
overbank area could increase the probability of outflanking when the water level (stream stage) is high. If stream stages can
be greater than or equal to the bank height, the spurs should be equal to the bank height. If flood stages are always lower
than the bank height, the spurs should be designed so that overtopping will not occur at the bank (DSCWM 2005).
Submerged spurs should have a height between 1/3 and 1/2 of the water depth (Jha et al. 2000).
Spur orientation relative to the river axis
Spurs can be attracting, deflecting, or repelling according to their inclination as shown in Figure 47. An attracting spur
points downstream and attracts the flow towards its head and thus to the bank, maintaining a deep current close to the
bank. A deflecting spur changes the direction of the flow without repelling it and creates a wake zone behind. A repelling
spur points upstream and diverts the flow away from itself. The first spur in a bend should always be attracting to minimize
the impact of the flow.
Spur shape
Spurs are basically bar shaped, but the end protruding into the water flow can be shaped differently (Figure 48). An
oval or elliptical spur, with the wider portion towards the bank, can change the hydraulic efficiency and reduce the direct
impact of the flood water on the spur body. Investigations have shown that the shape of the spur can affect the bed stress
distribution and the scour depth around a spur. For example, the extension of the high shear stress zone is smaller in
T-shaped spurs (Safarzadeh et al. 2010), whereas the maximum scour depth is less around L-shaped spurs (Hashemi Najafi
et al. 2008).
Spur length
When choosing the length of a spur, it is important to consider the safety of the opposite bank. If a spur is too long, it may
guide the river current during a flash flood to the opposite bank which will cause damage; if it is too short, it may cause
erosion of the near bank. As a general rule, the length of a spur should be no more than 1/5 the river width and no less
than 2.5 times the scour depth. Sometimes a spur is made long and strong with the aim of changing the river course by
repelling it towards the opposite bank, in which case the opposite bank should also be protected. Both the river width and
the width of the main flow channel to be deflected should be considered when designing the length of a spur (Jha et al.
2000)
The scour depth is given by
R = 1.35 (q2f )1/3,
where
R = the normal scour depth below high flood level (HFL),
q = discharge intensity in m3/s per metre width, and
f = Lacey’s silt factor, which depends upon the grain size of the river bed material (given for different materials in Table 14).
The value of q can be obtained from q =
Qf
in m³/s/m,
Water width
63
where the water width is the flood water width in the
river and
Table 14: Lacey’s silt factor for different
materials
Qf = 1.2Q – 1.24Q,
Type of material
where Q is the discharge.
Size of grains
(mm)
Silt factor (f )
Silt
Spacing
Very fine
0.052
The effect of a group of spurs depends on their length
Fine
0.081
Medium
0.158
and spacing. The spacing between two spurs depends on
Standard
0.323
the length of the spurs. The effect on flow is best fulfilled
if one strong eddy is created between each pair of spurs
Sand
(Figure 49). If the spacing is too wide, the effect of the
Medium
0.505
spurs will be insufficient as parts of the bank will remain
Coarse
0.725
unaffected. A spacing less than the optimum is wasteful
as it does not increase the effect. The length of bank
Gravel
protected by a spur is generally at least twice the length
Medium
7.28
of the spur projecting perpendicular to the river water
Heavy
26.1
current; thus spurs do not need to be closer than twice
Boulders
their projecting length. More exact calculations can be
Small
50.1
made using the formulae for eddy stability and energy
Medium
72.5
loss of river flow (HMGN 1990). In general, the spacing
Large
183.8
between two spurs should be 2–2.5 times the spur
Source: Varshney et al. 1983
length along a concave bank and 2.5–3 times the spur
length along a convex bank. In the case of a revetment
with spurs, the spacing can be increased
without causing harm to the bank (Jha et
Figure 49: Formation of eddy between two spurs
al. 2000).
Number of spurs along the stream bank
0.4
0.5
0.7
1.0
1.25
1.50
4.75
9.0
12.0
15.0
24.0
Groyne field
The number of spurs depends on the length
of stream bank to be protected and the
calculated space between spurs.
Launching apron
A launching apron should be constructed
to protect the spur from scouring at the
base. ‘Scouring’ is the name given to
Stagnation
the removal of the bed or bank of a
point
watercourse by the action of flowing.
When river training works are carried out
Source: HMGN 1990
to protect a river bank, the obstruction of
high flow discharge and the associated
changed water flow pattern can lead to
scouring in the form of a deep depression in the river bed close to the river training structure. Scouring can destabilize the
structure and thus measures need to be taken to counteract the effect. A launching apron is a flexible stone cover placed on
the bed of the river which settles into the scouring area as scouring takes place and covers the base and side of the scour
hole, preventing it from developing further. An example of an apron placed to protect a bank is shown diagrammatically in
the section below on guide banks.
Scouring is common around spurs and can result in their destruction. For an impermeable spur, the scour hole can reach
a depth of up to 1.0–1.7 times the design flow upstream of the pool, but it is much less for permeable spurs. A launching
apron should be constructed to protect impermeable spurs from scouring. For permeable spurs, it is sufficient to bury
material below the spur to fill the site of the potential scour hole.
64
Screen dams and beam dams
Figure 50: Bamboo porcupines used to form a spur
Screen dams and beam dams are sediment retention
structures. They are designed to trap medium to large
size debris and boulders carried downstream in flood
events in order to reduce the impact downstream.
This type of dam is often installed in alluvial fans,
along stretches with a steep slope, in wooded areas,
in areas with frequent mass movements, and along
narrow channel beds at the end of a valley just before
the stream or river enters an alluvial fan or plains
area.
The dams themselves must be constructed with strong
materials such as concrete or cement to withstand the
impact of heavy debris, whereas the retention portion
can be built with other materials. Other supporting
structures must also be constructed to protect the
banks and foundation. The dams require regular
maintenance; trapped debris and sediment should be
removed at regular intervals and after flood events to
maintain the storage capacity. These structures can
have a significant environmental impact depending
on their size and the materials used for construction.
The most common types of screen dams and beam
dams are the following.
Source: Rajendra Prasad Adhikary
Figure 51: RCC porcupines along a river bank (top)
and
combined with sandbags to form a bar projecting
into the Koshi river (bottom)
Screen dam with vertical bars. This type of dam is
mainly made to retain vegetative materials such as
tree trunks and branches. The vertical bars are usually
constructed from steel or concrete, although wood
can also be used in dams for small torrents and water
channels. The dam offers a high resistance to the
debris.
Beam dam with pylon bars, vertical opening, and
horizontal steel bars. These dams are constructed
with concrete and steel. The main purpose is to
control mass transportation of sediments that could
affect settlements or other infrastructure.
Porcupines
Porcupines are a form of permeable structure
designed to reduce flow and trap sediment. They
have pole-like projections in all directions, resembling
a porcupine with its quills sticking into the air. They
are used as flood control structures, and for river
bank and bed protection. Porcupines can be used
in a line forming a spur into a river, as silting aprons
for larger spurs, and in a longitudinal line along an
embankment. Originally such devices were made
of timber or bamboo (Figure 50), but these have a
limited lifespan. The use of wooden and bamboo
Source: Mitra Baral (top) and Shiva Kumar Sharma (bottom)
65
porcupines combined with vegetation to form a green wall is described in the chapter on bioengineering. This
section describes porcupines made of concrete.
There are two kinds of concrete porcupine in common use: reinforced cement (RCC; Figure 51) and pre-stressed
cement (PSC). Quality control of RCC struts is difficult because each strut cannot be tested separately, although a
rebound hammer can be used to test the uniformity of strength throughout. PSC porcupines are better in terms of
size, shape, strength, concrete mix, and steel used.
Porcupines can be constructed in two shapes, tetrahedral and prismatic (Box 17). The following are their main uses.
Bank protection as a bar. Porcupines can be used as a pro-siltation protection device for a natural river bank or
an embankment (Figure 54). The structures are flexible, which ensures stability against extreme water forces and
even earthquakes. Porcupines reduce the flow velocity, intercept and break eddies formed by floodwater, and fill up
scour holes with silt.
Porcupines are most commonly used as bars across, and aligned 2–5° upstream of, the flow. Each bar consists
of single, double, or triple rows based on the velocity of flow, the width of the bank line channel, and the spacing
between bars (Figure 54); the higher the velocity, the higher the number of rows. Single or double rows are used
when the bars are close to each other. The porcupines are generally placed so that they touch each other at the
base, and with the lines staggered if there are multiple rows. The bar extends from the highest flood level line to
Box 17: Porcupine design
Tetrahedral porcupines
The most common shape is tetrahedral. The porcupine is formed by assembling six concrete struts of the same length in
a tetrahedral pattern (Figure 52). Individual struts are bolted together, projecting beyond the joint. The bolts are passed
through holes made at the appropriate point using cheap polythene tubes during casting.
The size of a porcupine is denoted by the length of the individual struts, for example a 2 m porcupine or 3 m porcupine.
The struts of 2–3 m porcupines have a cross-section of 10 x 10 cm. The individual struts are at 60 degrees to each other,
thus a 2 m porcupine is about 1.7 m high and a 3 m porcupine 2.6 m high. The most common sizes in use are 2, 2.5, and
3 m. Mounted or long boom cranes are necessary to handle anything larger.
Prismatic porcupines
Prismatic porcupines are made with nine concrete struts joined in the form of a prism (Figure 53). Two end triangles are
formed first and then joined together with three struts placed at the vertices. Struts are bolted together as for tetrahedrons.
Figure 52: Tetrahedral porcupine (dimensions in metres,
Figure 53: Prismatic porcupine
shown here for a 3 m porcupine)
3.0000
0.4196
2.5981
0.4845
2.4000
3.0000
66
Figure 54: Plan (above) and cross-section (below) of a porcupine bar for embankment protection
Porcupine bars
Bank above flood water level
Prismatic porcupine on embankment slope
Tetrahedral porcupine on horizontal bank
Tetrahedral porcupine on river bed
High water level
Embankment
Sand char
Low water level
Bank
Trimmed bank
River bed
the deepest scour point of the bank line channel plus a marginal distance as a factor of safety. The bars may be
extended to link up with the continuous non submerged river bank or road – above the highest flood level – to avoid
out flanking. To reduce costs, the extension can be for an alternate or every third bar only.
Spacing between bars is not usually the same as for spurs. The main purpose of the bar is to obtain continuous
deposition of silt and not to deflect the flow away from the bank. However, wide spacing may create sand bars
around each bar leaving lagoons in between, thus closer spacing is preferred. When the curvature of the eroding
bank is sharp and the flow velocity is high, bars should be spaced less than 10 m apart.
Silting apron. Concrete porcupines are also used as silting aprons for spurs. Used in the form of a sunray, they
have shown encouraging results for filling scour pits around spurs.
Longitudinal Protection Structures
Longitudinal protection structures are installed on river banks parallel to the river course, generally with the aim of
protecting adjoining areas from inundation, erosion, and river meandering. They are usually constructed on natural
banks and extend for a considerable distance. The most common structures are embankments or levees in the form
of guide bunds or banks, afflux bunds, and approach embankments. Very often, spurs are constructed together with
longitudinal structures to protect the latter. Some common longitudinal structures are described in the following.
Levees or earth fill embankments
Levees, or marginal embankments, are dam-like earthen structures constructed along a river in order to protect the
surrounding countryside from flooding and/or to confine the course of a river to provide higher and faster water
flow (Figures 55 and 56). They are usually constructed for long stretches along a river in low lying areas with an
extended floodplain (Figure 57).
67
Figure 55: Levee along the Rapti River in Nepal
Figure 57: Levees (marginal embankments) protected
by spurs along a river in the plains
River
Spur
Spurs
Spur
Marginal
embankment
Spur
Figure 56: Extension of flooded river before and after
levee construction
Spur
Before
Road and rail
bridge
Flood plain
River
Flood plain
Spur
After
Levee
Levee
Spur
Spur
Restricted width of river
Levees are usually constructed by piling earth on a cleared level surface. The type of fill material used for
construction usually depends on the materials available in the local area. The levee must be designed and
constructed very carefully as failure can result in catastrophic impacts.
Both sides of the levee should be properly constructed. The slope is fixed to ensure stability, and ultimately
depends on the material that the levee is made of and its height. The sides should be strengthened with riprap
(see below) to prevent erosion.
The slopes of the upstream and downstream faces of the embankment should be flat enough to provide sufficient
width at the base to ensure that the maximum shear stress under flood conditions will remain well below the
corresponding maximum shear strength of the soil, in order to provide a suitable factor of safety.
Specific design criteria for levees are given in Box 18.
Guide banks and other approach embankments
Guide banks are structures built to guide a stream or river through a bridge opening or towards other hydraulic
structures such as weirs, especially when river flow level is markedly higher than usual. The aim is to confine the
68
Chapter 6: Physical Methods for River Training
Box 18: Design criteria for levees
Freeboard
The minimum vertical distance between the maximum flood level and the top of the levee (the crown or crest) is generally
taken to be 1.5 times the height of the wave (hw), which is calculated from the following:
hw = 0.032√VF + 0.763 - 0.271 (F)¼
or
hw = 0.032√VF (for F > 32 km),
(for F < 32 km)
where
V = velocity of wind km/hr, and
F = straight length of water surface in km.
Width
The top width of the embankment should be sufficient to keep the seepage line well within the levee. For a small levee, this
top width is generally governed by the minimum roadway width requirements.
The minimum top width (A) of an earthen levee can be calculated as follows:
A = H/5 + 3
A = 0.55√H + 0.2H
A = 1.65 (H + 1.5)1/3
for a very low levee,
for a levee lower than 30 m, or
for a levee higher than 30 m,
where H is the height of the levee.
river within a reasonable waterway and direct the flow in a manner that ensures its safe and expeditious passage
(Varshney et al. 1983). They also reduce or eliminate local scour at the embankment and adjacent piers (Julien
2002). In a wide river lined by levees, a series of diversion structures may be used to guide and narrow the water
course and protect the levee or highway embankment, where a highway or other bridge crosses the river. These
consist of an afflux embankment or bund, an approach embankment, and the guide banks themselves (Figure 58).
Guide banks are constructed in a river in order to:
confine the flow to a single channel,
improve the distribution of discharge across the width of a river thus controlling the angle of attack by a flash
flood,
protect weirs, barrages, or other hydraulic structures constructed in the river such as intakes from flash floods,
control the meander pattern of a river,
control overtopping of natural embankments in a flash flood and protect adjacent land from flooding,
reduce erosion of banks by the water current,
prevent sliding of soil as a result of the draw down effect of the flood water level,
facilitate smooth transportation of water, and
prevent piping of water through the banks.
Two guide banks are generally required when the waterway opening is in the middle of a wide flood plain or is
a braided stream where the direction of the main flow can shift from side to side. A single guide bank may be
sufficient at a location where the river is confined to one side of a valley and it is possible to take advantage of a
natural non-erodible bank such as a hard rock exposed surface. It is essential to check the load bearing capacity of
the river bed sub-soil before choosing the location of a guide bank. The minimum width between the guide banks
should be sufficient to provide the required waterway opening during the anticipated flash flood discharge.
The design of guide banks is described in Box 19.
69
Figure 58: Guide banks and other approach embankments
Afflux embankment
River flow
Afflux embankment
Guide banks
Launcing apron
Launcing apron
Approach embankment
Approach embankment
Bridge
Guide banks
Box 19: Design of guide banks
Length
For shifting alluvial rivers, the length depends on the distance necessary to secure a straight run for the river, and the
distance necessary to prevent the formation of a bend in the river so as to avoid the angle of attack of the anticipated flash
flood.
Plan shape
Ideally, the guide bank should have a converging curved shape forming a bell mouth entry to the waterway. The axis
should be parallel to the principle direction of flood flow through the opening. This shape is particularly suitable where
the direction of flow can vary. In most cases, the main sections of the two banks are constructed parallel to each other, but
other forms are possible, for example curved or converging.
Embankment section
The angle of the embankment slope is calculated according to the subsoil conditions, the angle of repose of the
embankment material, and the type of slope revetment provided (see below): the slope should usually have a vertical to
horizontal ratio of between 2:1 and 3:1 (Singh 1980). In general, the top of the embankment is made wide enough to
accommodate vehicles for construction and maintenance purposes. Guide banks should normally extend above the design
high water level with a freeboard allowance of 1–1.5 m depending upon the discharge condition (Singh 1980). Lower
guide banks that can be overtopped under high flood discharge condition may be preferred in some cases. Under these
conditions, the top of the bank and outside slope must be protected against erosion.
Spacing between the guide banks
The layout of the guide banks should be such as to guide the flood smoothly throughout the guide bank length. Generally,
the guide banks are constructed to form a symmetrical pair. They should confine the river within a reasonable channel that
can ensure safe and rapid passage of water during a flash flood. The confined width of the river between the guide banks
in an alluvial river can be calculated using Lacey’s formula (Singh 1980):
70
L = 4.75 Q½,
where
L = constrained width of the river in m, and
Q = maximum discharge in m³/s of the river during a flash flood.
This equation is for finding the wetted perimeter. In practical cases, the width is slightly more than the wetted perimeter, and
the formula is modified to:
L = 5 Q½.
The calculated waterway should be multiplied by 3–6 to give the spacing between the embankments; the exact number
depends on the width, nature, river characteristics, and level of protection to be provided (Jha et al. 2000).
Pitching
The inside slope of the embankment is subjected to erosion from the river flow, particularly during floods and flash floods.
The continuous movement of water saturates the embankment material as a result of pore water pressure. Sudden increases
and decreases in the water level can change the water inflow and outflow in the embankment material and damage the
embankment. Hence, the inside slope should be protected by stone pitching. The usual thickness of the pitching varies from
40–60 cm. The thickness can be determined from the formula t = 0.60 Q1/3, where t is the thickness in metres and Q is the
maximum river water discharge in m3/s (Varshney et al. 1983).
Launching apron
Stone pitching protects the face of the bank. However, floods can induce scouring at the toe which would undermine the
pitching and cause its collapse. To prevent this, a stone cover or launching apron is laid beyond the toe of the bank on the
horizontal river bed (Figure 59). As the scour undermines the apron starting at its farther end and working back towards the
slope, the apron falls to cover the face of the scour, with the stones forming a continuous carpet below the permanent slope
of the guide bank. The apron must have sufficient stone to ensure complete protection of the whole of the scour face. The
length of the scoured face is equal to 5√D, where D is the anticipated scour depth below the apron.
The scouring effect is a function of the gradation of the silt available in the river bed and the discharge of the flowing water.
It can be calculated using the following formula (Varshney et al. 1983):
Lacey’s silt factor is:
f = 1.76 mr½,
where mr is the average diameter of the river bed material.
Depth of scour (R) is given by
R = 0.47 (Q/f )1/3,
where Q = is the maximum discharge in m³/s of the river during a flash flood.
Figure 59: Cross-section through guide bank (numbers indicate relative values for any given size
4.0
Stone launching apron
2
2
1.5H
1
2 layers thick
1
Concrete slope revetment
Original bed
Launched
position
after scour
1
2
Estimated depth of scour, H
Source: HMGN 1990
71
Concrete embankments
Concrete embankments are made from cemented bricks, stones, or concrete. These are thin but strong
embankments usually installed in urban reaches of water courses where there is not enough space to build
more massive structures. They can also be combined with earth fill structures. The construction cost of concrete
embankments is higher than that of earth fill embankments and such an embankment has a significant impact on
the environment and often destroys the ecology of riparian areas.
Revetments and rock riprap
Revetment refers to a continuous artificial surface on a river bank or embankment slope and part of the river bed
which is designed to absorb the energy of the incoming water and protect against erosion by the river current.
Revetments are usually placed along the concave side of a river bend where river velocities are high. Upstream
from barrages, revetments may be used to hold approaching river banks in their existing positions. Revetments can
be flexible or rigid. They can be constructed from various materials including rock, stones, stone-filled gabions,
concrete slabs, timber piles, bamboo piles, old tyres, and sandbags.
If there is a potential for scour at the toe, the revetment must be extended down to the expected level of the scour
and sufficient material added in the form of a thickened toe or horizontal apron such that the toe material will
launch to a stable slope as the bed scour develops.
Rock riprap is one of the most common types of material used to make a revetment (see, for example, FHWA 1989;
MDEP 1997). Riprap is a layer or facing of loose rough rock or other material used to armour embankments,
streambeds, bridge abutments, pilings, and other structures subject to erosion by wave action and impact damage
from debris carried by the waves. Riprap absorbs and deflects the energy of floodwater and debris before it reaches
the defended structure, while the gaps between the rocks trap and slow the flow of water. Riprap has structural
flexibility and can be constructed with locally available materials.
A riprap revetment is composed of three sections: an armour or stone layer, an underlying filter layer, and a toe
protection layer. Armour is the outer layer of rough angular rock. The underlying filter layer supports the stone
against settlement, allows groundwater to drain through the structure, and prevents the soil beneath from being
washed through the armour by waves or groundwater seepage. The toe protection prevents downward movement
of the riprap and is constructed by trenching in the riprap at the toe of the slope. Figure 60 shows the crosssection through a typical riprap revetment. Various designs can be used for toe protection depending on the exact
requirements and physical location (Figure 61). The design considerations for a rock riprap revetment are described
in Box 20. More detailed information can be found in FHWA (1989). This approach can be adapted for other types
of revetment.
Apart from rock riprap, reformed flexible ‘mattresses’ can also be used to form revetments. They can be prepared
from a variety of materials including willow and timber, reinforced asphalt, soil, cement, and articulated concrete.
The design and construction method depend on the type. Reinforced asphalt mattresses are the most durable and
abrasion resistant of the above types but highly mechanized techniques are required for their manufacture and
placing. Gabion mattresses consist of flat mattresses fabricated from wire mesh and filled with stone. Mattresses are
generally used to protect a bank against high velocities where stone sizes are very small.
Bags filled with sand or concrete can also be used as revetment. This type of protection is used where gravel is
available but large stones are rare. This protection is rigid and almost all the failures are due to undermining of the
toe at the ends of the protection.
Porcupines used as embankment protection
Transversal arrangements of porcupines are used to reduce water velocity and encourage siltation, as described in
the first part of this chapter. Longitudinal arrays of porcupines can also be used as a pro-siltation device to stabilize
a river bank (Figure 63). The porcupines can be made of timber or bamboo as well as concrete. The bamboo and
timber structures have a shorter life than concrete porcupines but can be more environmentally acceptable. They
72
Figure 60: Cross-section of typical rock riprap revetment
Access for
site inspection
Existing bank line
2
1
Flood construction level
1
Thickness
1.5xD50 or = D100
2
Scour protection
(Riprap filled toe trench)
Prepared bank grade
Existing bed level
Gravel filter layer
Maximum scour depth
Launched scour
protection
Figure 61: Different designs for the toe of a rock riprap revetment
T
T
a T
b=2T
(a) Toe inclined
c b
(b) Toe apron
T
T
c = 5T
a = 3T
a = 1.5T
c = 5T
(c) Toe apron on bed
(d) Thick slope
T
Legend
Riprap
Filter layer as required
Bedrock
Water level
(e) Inclined slope on bedrock
73
Box 20: Design considerations for rock riprap revetments
Rock size
Riprap is classified as either graded or uniform. Graded riprap contains a mixture of stones which vary in size; uniform
riprap contains stones which are similar in size. For most applications, graded riprap is preferred to uniform riprap.
Graded riprap forms a fexible self-healing cover, while uniform riprap is more rigid and cannot withstand movement of the
stones. Graded riprap is cheaper to install, requiring only that the stones be dumped so that they remain in a well-graded
mass (MDEP 1997). Graded riprap is described in terms of the median diameter D50: in a mixture of stones, 50% of the
rock by weight will have a diameter above D50 and 50% below. The largest stones should not exceed 1.5 times the D50
specified.
The stability of the structure is provided by the well
graded rock layer and the average rock diameter. Rock
riprap on a stream bank is affected by the hydrodynamic
drag and lift forces created by the velocity of flow
past the rock. These effects are resisted by the force
components resulting from the submerged weight of the
rock and its geometry. The required D50 is determined
from the expected maximum velocity of flow under flood
conditions as shown in Table 15.
Rock shape
Table 15: Determining median diameter (D50)
of riprap from maximum flow velocity
Maximum flow velocity
(cfs)
Riprap D50
cm
inches
16
91
36
13
61
24
11
46
18
10
38
15
8
25
10
Stones should be shaped so that the least dimension
6
15
6
of the stone fragment is not less than one-third of the
4
7.5
3
greatest dimension of the fragment. Flat rocks should not
Source: MDEP 2003
be used for riprap. Blocky and angular shaped rocks
with sharp clean edges and relatively flat faces are
good. If rounded stones are used, they should be placed on flatter slopes (not exceeding 2.5:1 horizontal to vertical) and
the recommended median rock diameter should be increased by 25% with a comparable increase in the thickness of the
revetment (USACE 1991).
Thickness
The minimum thickness of the riprap layer should be 2.2 times the maximum stone diameter for a D50 of 30 cm or less, but
not less than 15 cm, and twice the D50 for a specified D50 greater than 30 cm.
Figure 63: Bamboo porcupines as protection along an
embankment
can be intertwined with living plants to form a stable
structure as silt builds up and the plants root into the
ground (see Chapter 4).
Other Protection Structures
Sandbagging
Sandbags can be used to reinforce structures (e.g.,
Figure 51) and to build (emergency) dikes (Hellevang
2011). They are widely used to control or reduce
the devastating effects of floods particularly in plains
areas. Figure 64 shows a typical use of sandbags in a
toe wall along an embankment.
74
Sandbags can also be stacked to make a barrier
against rising flood water as well as in areas where
flash floods are likely (Figure 65). The sandbag wall or
barrier should be constructed on a firm flat surface to
prevent seepage. A trench can be dug and the bottom
Filter blanket
A geo-textile or stone filter blanket should be placed under the riprap to prevent water from removing the underlying soil
material through voids in the riprap. Generally, the filter blanket is made from a layer of well-graded gravel or sand-gravel,
or synthetic filter fabric materials. The design of a gravel filter blanket is based on the ratio of the particle size in the
overlying filter material to that of the base material in accordance with the following criteria (Brooks 1989, cited in PBC
2000):
D15c/D85f <5 <D15c/D15f <40,
where D15 and D85 refer to the 15% and 85% sieve passing sizes, and subscripts ‘c’ and ‘f’ refer to the ‘coarse’ and ‘finer’
layers, respectively.
Geo textiles can also be used as filters in place of or in combination with gravel filters. They are both cheaper and easier to
install.
Toe design and scour depth
The revetment toe must be protected from undermining by scour. A deep scour hole can form at the tip of the revetment
where the flow velocity is much higher than the average channel velocity (Figure 62). The scour hole can undermine the
bank leading to a collapse of the whole structure. An apron can be constructed to fill the scour hole (Figure 59).
This method is recommended for cohesionless channel beds in which deep scour is expected. In cohesive channel beds, the
bank revetment should be continued down to the expected worst scour level and the excavated area refilled as shown in
Figure 59 (Julien 2002).
Figure 62: Scour hole and expected scour
Design high water
T = Thickness of revetment
H = Thickness of apron
Projected length
Apron
T
H = 2.5T
Expected scour
Scour hole
layer of bags placed in it to improve stability. Plastic
sheets can be used to help seal the dike if available. It
is important to construct the barrier properly to ensure
that the result is effective; Hellevang (2011) provides
detailed instructions. Bags can be made from various
materials and in different sizes but woven polypropylene
is the most common. Bags with a filled weight of no
more than 30–40 pounds (14–18 kg) are easier to
handle. Sand is the easiest and most available material
for filling and shaping the bags. Silt and clay can also
be used, but working with these materials is more
difficult.
Figure 64: Toe wall constructed from bamboo and sandbags
Source: R Adhikary
Whatever the final use, it is important to fill the bags
properly. The bags should be about one-half full and
tied near the top so that the sand can move easily
(Figure 66). Overfilled bags and bags tied too low lead
to gaps in the dike or wall, which allows water to seep
through.
Source: Mercy Corps
75
Figure 65: Sandbag dike
Width: 3 times height
Height
Bonding trench 1 sack deep, 2
sacks wide
Water side
Strip sod before laying bottom
layer, if possible
Figure 66: Correct way to fill sandbags
YES! Fill the sandbag
about one-half full
and tie near the top,
if it needs to be tied
Figure 67: Concrete channel lining
NO! Sandbag
filled too full
NO! Sandbag
tied too low
Source: Hellevang 2011
76
Source: Nabin Baral
Channel lining
Channel lining is a protective layer used to protect the banks and bed of a watercourse against erosion. Channel
lining can help increase the velocity of flow to ensure easy transport of sediment and reduce deposition in the
channel bed. It is recommended in catchments highly prone to erosion, particularly in urban and alluvial fan
reaches. However, channel lining can have a marked environmental impact and the necessity and the type of
structure should be carefully assessed.
Channel lining structures can be made from many materials including concrete (Figure 67), gabions, and wood, as
well as earth, rocks, asphalt, and plastic. Concrete and cement linings have a higher environmental impact, natural
materials generally a lower one. Wood channel linings are usually cheaper to install and maintain than those made
of other materials.
Bamboo piles
Bamboo can be used in the form of piles to strengthen a foundation or stabilize a flood embankment or river bed.
The rows of bamboo piles should be firmly fixed with a rope or iron wire. Piling in wet soil is very easy but may
otherwise require more strength. It may be necessary to excavate small holes in boulder covered parts of the river
bed. Two parallel rows of piles can be prepared and the space between them filled with boulders and pebbles as a
toe protection measure for flood embankments (Box 21).
Box 21: How to use bamboo piles to develop a protection structure
Materials
B
amboo piles
D
igging tools, hammer
Boulders or pebbles
Installation
1. Drive piles into the ground at least 1 m deep by hammering. The piles should be about 40 cm apart and driven in to
leave about 1–1.5 m exposed at the top.
2. Where there are boulders, excavate a small area and hammer the pile in. Fill in around the pile.
3. Tie the piles together with rope or iron wire.
4. Fill the space between parallel rows of piles with boulders and pebbles as a toe protection measure.
77
Chapter 7: Structural Measures for
Flood Management in the Context
of Integrated Water Resources
Management
The basis of integrated water resources management (IWRM) is that different uses of water are interdependent.
Managers, whether in the government or private sectors, have to make difficult decisions on water allocation.
More and more they have to apportion diminishing supplies between ever-increasing demands. Drivers such as
demographic and climatic changes further increase the stress on water resources. The traditional fragmented
approach is no longer viable and a more holistic approach to water management is essential (UN-Water 2008).
In the words of the Global Water Partnership: “IWRM is a process which promotes the coordinated development
and management of water, land, and related resources, in order to maximize economic and social welfare in an
equitable manner without compromising the sustainability of vital ecosystems and the environment. ... It is a crosssectoral policy approach, designed to replace the traditional, fragmented sectoral approach to water resources and
management that has led to poor services and unsustainable resource use. IWRM is based on the understanding
that water resources are an integral component of the ecosystem, a natural resource, and a social and economic
good” (GWP 2012).
The concept of IWRM offers solutions to the water crisis by linking water to other vital resources in a holistic
manner and viewing the whole water cycle together with human interventions as the basis for sustainable water
management. The specific approaches used need to be tailored to the individual circumstances of a country or local
region.
Integrated Flood Management
Sustainable and effective management of water resources demands a holistic approach, linking social and
economic development with the protection of natural ecosystems and appropriate management links between
land and water uses. Therefore water-related disasters such as floods and droughts that play an important part in
determining sustainable development also need to be integrated within water resources management.
Flood management plays an important role in protecting people and their socioeconomic development in
floodplains. Unfortunately, strategies that rely on structural solutions (dams and reservoirs, embankments, bypass
channels) alter the natural environment of the river, resulting in losses of habitat, biological diversity, and productivity
of natural systems. The need for sustainable development has highlighted the importance of addressing the negative
consequences on the environment of these flood-protection measures. Environmental degradation has the potential
to threaten human society in the areas of safety of life, economic wellbeing, and food and health security. Thus it is
essential to consider environmental impacts in flood management activities (WMO 2006b).
There are no universal solutions that determine environmentally-friendly flood management practices. Practices
should be adopted that suit the particular circumstances of a basin using the three-way approach of avoiding,
reducing, and mitigating adverse environmental impacts without compromising the flood management objective.
This, in essence, is integrated flood management (IFM), a process that promotes an integrated – rather than
fragmented – approach. IFM integrates land and water resources development in a flood plain within the context of
IWRM, and aims to maximize the net benefits from flood plains while conserving the environment and minimizing
78
Chapter 7: Structural Measures for Flood Management
loss to life, infrastructure, and property from flooding (APFM 2004; WMO 2006a). An integrated approach means
considering all the impacts in a holistic manner by looking at linkages and interdependencies between upstream
and downstream areas as well as between the river course and the flood plain. This means looking at the whole
river basin with its natural boundaries in line with the flow patterns of water, rather than individual administrative
areas.
IFM uses a multidisciplinary approach to flood management and involves a wide range of stakeholders,
including professionals from different agencies and fields directly or indirectly related to flood management, and
representatives of those most likely to be affected by flooding, as well as by anti-flooding measures. This chapter
describes some integrated basin wise approaches to flood management, using some of the techniques described in
the previous chapters and taking into account human and environmental needs.
Passive Flood Control Measures
Passive flood control measures refer to the avoidance of activities that accelerate flood discharge. The main purpose
is to favour natural flow retention through the interdisciplinary management of water and land resources along
watercourses and preserving and improving the general environmental conditions. The main control measures are
as follows (Colombo et al. 2002: 29–30):
adaptation of cultivation in the neighbourhood of watercourses to flooding, taking into account the resistance
and susceptibility to damage of different crops;
development of forms of cultivation suitable for the local conditions in river plains;
transfer of local cultivation to safer areas;
public acquisition of land and structures that are more frequently subjected to floods;
safeguarding of available lowland run-off zones covered by woods, coppice, and grassland;
cessation of construction in flood plain areas;
restoration of retention areas through withdrawal and displacement of dams as well as reactivation of old arms;
preservation of natural-like trenches, and environmentally-friendly development of already trained water systems;
promotion of land suitable for water retention or needed for building flood protection structures in land use
plans;
setting up integral land consolidation procedures both to favour passive control and foster sustainable
agricultural practices; and
building or preservation of structures to slow runoff.
Flood Storage Reservoirs
Flood storage reservoirs or basins control floods or flash floods by detaining and storing water and then releasing
it slowly at a controlled rate, thus reducing the destructive flood peak. In this method, effective flood mitigation is
obtained by having an adequately-sized reservoir or group of reservoirs immediately upstream from the area to be
protected. The most effective reservoirs for flood control are those located in a broad floodplain. However, these
require the construction of a very long dam or embankment bounding the area where the flood water is to be
stored, and a large area of valuable agricultural land would have to be flooded. Such a reservoir would catch most
of the run-off water that would cause flooding to the nearby area. Sites further upstream require smaller dams and
less valuable land, but they are less effective in reducing flood peaks (ADPC/UNDP 2005).
The terms ‘detention’ and ‘retention’ basins are sometimes used interchangeably, but there are some differences
between them.
A detention basin is an area where excess storm water is stored or held for a short time before it drains out to a
natural watercourse, or for a longer time for agricultural, consumptive, aesthetic, recreational or other uses. The
water retention also helps reduce the amount of pollutants transported by runoff (Colombo et al. 2002). Detention
basins are sometimes called dry ponds or holding ponds because they eventually empty out to downstream.
In some cases water remains in the basin almost indefinitely, reduced only by the net amount lost by evaporation
and absorption at the surface. Such basins are called retention basins, wet ponds, or wet detention basins.
79
The design of water detention or retention basins includes a number of factors such as the size, stability, and
impermeability of the dam or embankment, the design of the elevation and capacity of the outlet for releasing the
water, and the storage volume required. Diversions are often required to channel the flow from the main river to the
basin area. A variety of engineering work is required to install flood storage reservoirs.
Some of the specific considerations in the planning, design, operation, and maintenance of water retention basins
have been summarized by Colombo et al. (2002):
Hydrological elements such as the basin capacity and spillway should not be undersized. However, large
retention facilities might have a high ecological impact.
Correct hydraulic dimensioning of the operation facilities is needed similar to that for conventional dams,
including correct design of the bottom outlet, the spillway, and the bypass.
The geological setting must be taken into account, including the geotechnical properties of soils and
groundwater conditions, especially for permanent lakes.
Ecological aspects should be considered, for example the ecological continuity of the flowing water, sediment
accumulation, and channel shaping downstream of the basin. The bottom outlet should be designed so as
to minimize interruption of the ecological continuum of the water flow. Wherever possible, the construction of
retention installations should be accompanied by ecological compensating measures. Limnological development
and successional use has to be analysed if a permanent lake is to be planned.
Bioengineering criteria should be adopted that favour restoration of the natural condition of the river course and
thus the increase of biodiversity within the riparian habitat, especially in urban areas.
The commissioning, operation, and maintenance phases should consider safety of the population in connection
with urban and industrial development, land reclamation, and new traffic routes, which usually emerge in the
vicinity of retention basins.
Areas subject to flooding during events with a return period of five years or less should not be used as arable
land. At best, they can be used as grassland.
Forest management should take into account the resistance of the tree cover against floods as well as its fitness
for the site. Storage of timber in the retention area should be forbidden.
Recreational facilities must either be positioned outside the retention area or be designed to withstand floods, or
appropriate safety measures should be taken.
Use of basins for power generation, water replenishment, and/or storage of drinking and service water must be
secondary to the use as a flood retention basin. In general, any secondary use of such basins should have no
noticeable influence on the main purpose.
If provision for habitats and nature conservation is made in the basins, natural succession should be the prime
objective. Again, this use must be secondary to flood reduction.
Existing retention basins should be kept in good condition and adjusted to current engineering standards based
on revised hydrological and slope stability conditions.
A qualified inspector should be appointed for major retention basins and an overseer (e.g., a water licensee) for
other basins.
The earth dam surfaces should be protected against erosion and infiltration by surface water. A dense vegetation
cover such as turf can be effective.
Dams or embankments, emergency spillways, and outlets should be checked periodically (and always after each
storm) for erosion damage, piping, settling, seepage, or slumping along the toe or around the barrel. Regular
maintenance should be carried out. Maintenance on the outside of the dam includes removing debris from
around the inlet, mowing the embankment to keep bushes and trees from becoming established, checking the
outlet pipe for cracks or other damage, and removing rodents that burrow into the dam.
Sediments within the basin must be periodically removed and disposed of in order to maintain the volume
for storage of flood water. As a rule of thumb, accumulations over one-half the design volume should not be
allowed.
The longitudinal structures enclosing the basin must be planned or adapted with maintenance roads in order to
guarantee accessibility to people and machines.
In addition to the general recommendations above, there are some specific recommendations for lateral retention
basins (i.e., basins constructed on flat areas close to the water course).
80
Chapter 7: Structural Measures for Flood Management
Locate in central or lower portions of the catchment, where the existence of flat areas enables identification of
large surfaces for the storage of significant volumes of flood water.
Land planning of river catchments should identify bonds and prescriptions for preservation of areas that can
potentially be used for lateral retention basins.
The entrance and exit points between the river course and the basin must have protective structures installed in
order to avoid erosion.
Construction of artificial lakes and islands, and plantation of small wooded areas may favour the colonization of
many different plant species.
River Corridor Rehabilitation and Restoration
River corridor enhancement, rehabilitation, and restoration is a concept that involves different activities with a
degraded river or stream in order to return it to its natural condition. In practice, restoring a river or stream to its
original state is almost impossible. However, some limited enhancement works can be done to restore the river
corridors.
The general environment of the riparian zone and the ecological role of the riparian vegetation must be taken into
account in interventions on fluvial systems. In addition to its ecological function for aquatic habitats and terrestrial
wildlife, this vegetation provides important socioeconomic benefits (e.g., flood defence, scenic and aesthetic
quality, and leisure) as well as regulation of ecosystems (e.g., riverbed stability, erosion control, filtering/retention of
sediments, flood defence, wastewater treatment, and pesticide control).
Ecologically oriented interventions in the riparian vegetation should consider the following criteria (Colombo et al.
2002: 32):
keeping a structure of vegetation of different ages that allows the presence of both shrub and tree layers;
periodic cutting and selective thinning of adult trees that present problems of stability, and elimination of
invading species in favour of autochthonous species and, possibly, valued species; and
keeping shrubby vegetation where possible, since it can bend easily during floods and does not obstruct bridge
sections.
The influence of periodical flooding processes, sedimentation, and erosion can be felt in the transition areas of
riparian vegetation. Structural measures to regulate the river, such as large hydraulic infrastructures that aim to
improve water flow conditions, usually cause environmental changes throughout the fluvial ecosystem. These can
be reflected in the destruction of habitats and vegetation, alteration of physico-chemical water characteristics, and
other modifications at the level of the whole ecosystem.
The overall interventions on river corridors should focus on maintenance, rehabilitation, and restoration of wellstructured corridors consisting of wide lateral strips containing varied topographical elements and populated by
autochthonous grass, shrub, and tree species.
Table 16 summarizes some of the main measures used to enhance, rehabilitate, and restore river corridors. The most
common efforts focus on enhancing riparian zones by planting grasses, bushes, and trees; stream bank stabilization;
the removal of dams and other man-made structures; and stocking the river with fish or other living organisms.
The following principles related to the ecology of river systems can help practitioners when planning and
undertaking rehabilitation projects (Cottingham et al. 2005).
Riverine ecosystems are structured hierarchically, with important processes operating at a range of spatial scales,
from large regional and catchment scales, to sub-catchment and reach scales, and ultimately down to smaller
site and micro-habitat scales.
Riverine ecosystems can also be highly dynamic and variable in space and time. As such, stream ecosystems
in good condition are resilient to periodic natural ecological disturbances, such as droughts, floods, and fires,
which can help drive important physical and biological processes.
Hydrological connectivity provides strong spatial connections along river networks and between rivers and their
floodplains, and plays a key role in ecological processes such as nutrient and energy cycling (spiralling), and the
recovery of populations and communities following natural and human induced disturbance.
81
Table 16: Measures to enhance, rehabilitate and restore river corridors
Riverbed
Margin
Flooding area
Cleaning and removal of obstructions
Cleaning and removal of obstructions,
managing damaged trees
Cleaning and removal of obstructions
Recovery and restoring of natural
conditions
conditions
conditions
Ecological and aesthetic valorization
River-bed modification
Revegetation planting and seeding
Revegetation planting and seeding
Meandering
Stabilization, protection, and/or
natural, semi-natural, and artificial
revetments
Increasing hydrological communication
with the river bed and margins
Narrowing/widening
Plaiteds, reed rhizomes
Level lowering
Ecological flow regime
Gabions, geo-textiles, foundationstones, and similar, used with or
without vegetal materials
Humid zones and increase of habitat
diversity
Substratum modification
Flow deflectors
Flood retention basins
Silt traps
Slope modelling
Compartmentalization systems
Alternative river-beds
Buffer strips
Source: Colombo et al. 2002: 35
Stream rehabilitation activities are embedded in the hierarchical organization mentioned above, and should
begin with an examination of large-scale factors that might constrain processes acting at smaller spatial scales.
Longitudinal processes should also be considered, as the source of degradation can be some distance from
the locations where ecosystem impacts are evident and rehabilitation is proposed, and loss of connectivity may
constrain the biotic response to physical changes in the channel.
The most effective form of rehabilitation is to prevent degradation of river ecosystems in the first place. The
highest priority should go to protecting the remaining high quality river systems (or parts thereof), particularly
those that serve as important refugia and are a potential source of colonizing organisms.
Rehabilitation should aim to increase the resilience of river ecosystems to natural (and further human-induced)
disturbances so those ecosystems become self-sustaining and capable of responding to large-scale processes
such as climate change and the condition of catchments.
Where possible, rehabilitation efforts should aim to work with natural processes. This means considering
rehabilitation at broader scales than is often practised and choosing realistic rehabilitation targets. Given the
nature of human impacts, it is unlikely that degraded streams can be returned to their pre-disturbance condition.
In such circumstances it can be inappropriate to adopt ‘return to natural’ as the target for rehabilitation.
The fragmentation of populations, coupled with sometimes low levels of connectivity (whether because of human
interventions such as barriers or naturally poor dispersal abilities), means that many plant and animal species will
respond to habitat restoration only very slowly. Isolation may therefore be a major constraint to biotic recovery,
and in some cases this could take years or decades. Further, local extinctions may preclude full population
recovery.
For rehabilitation to be successful, proper planning and interaction with stakeholders must complement the
ecosystem perspective listed above. Socioeconomic considerations will play a large role in determining where and
when different rehabilitation measures are applied.
82
References
References
ADPC/UNDP (2005) A primer for integrated flood risk management in Asia. Bangkok, Thailand: Asian Disaster Preparedness
Center (ADPC); United Nations Development Programme (UNDP)
Allen, HH; Leech, JR (1997) Bioengineering for streambank erosion control: Report 1 guidelines, Technical Report EL-97-8.
Vicksburg (MS), USA: U.S. Army Corps of Engineers, Waterways Experiment Station
APFM (2004) Integrated flood management: Concept paper. Geneva, Switzerland: Associated Programme on Flood
Management. www.unwater.org/downloads/concept_paper_e.pdf (accessed 24 April 2012)
APFM (2007) Guidance of flash flood management: Recent experiences from Central and Eastern Europe. Geneva, Switzerland:
Associated Programme on Flood Management
APROSC (1991) Glossary of some important plant and animal names in Nepal. Kathmandu, Nepal: Agricultural Projects Services
Centre
Bajracharya, SR; Mool, PK; Shrestha, BR (2007) Impact of climate change on Himalayan glaciers and glacial lakes: Case studies
on GLOF and associated hazards in Nepal and Bhutan. Kathmandu, Nepal: ICIMOD
Bentrup, G; Hoag, JC (1998) The practical streambank bioengineering guide: User’s guide for natural streambank stabilization
techniques in the arid and semi-arid Great Basin and Intermountain West. Aberdeen (ID), USA: USDA-Natural Resources
Conservation Service, Plant Materials Center
Bentrup, G (2008) Conservation buffers: Design guidelines for buffers, corridors, and greenways, General Technical Report SRS109. Asheville, North Carolina: Department of Agriculture, Forest Service, Southern Research Station
Brien, H; De Blasio, FV; Elverhoi, A; Hoeg, K (2008) ‘Erosion and morphology of a debris flow caused by a glacial lake outburst
flood, Western Norway.’ Landslides 5: 271–280
Brooks, WE; Willett, JC; Kent, JD; Vasquez, V; Rosales, T (2005) ‘The Muralla Pircada – an ancient Andean debris flow retention
dam, Santa Rita B archaeological site, Chao Valley, Northern Peru.’ Landslides 2: 117–275
Brown, SA; Clyde, ES (1989) Design of riprap revetment, report FHWA-IP-89-016, prepared by Sutron Corporation for U.S.
Federal Highway Administration, Washington DC, USA
Cenderelli, DA; Wohl, EE (2001) ‘Peak discharge estimates of glacial-lake outburst floods and “normal” climatic floods in the
Mount Everest region, Nepal.’ Geomorphology 40: 57–90
Colombo, AG; Hervas, J; Arellano, ALV (2002) Guidelines on flash floods prevention and mitigation. Ispra, Italy: NEDIES http://
eusoils.jrc.ec.europa.eu/Esdb_Archive/eusoils_docs/other/EUR20386EN.pdf (accessed 19 June 2012)
Cottingham, P; Bond, N; Lake, PS; Arthington, A; Outhet, D (2005) Recent lessons of river rehabilitation in eastern Australia,
Technical report. Canberra, Australia: CRC for Freshwater Ecology
Cruden, DM (1991) ‘A simple definition of a landslide.’ Bulletin of the International Association of Engineering Geology 43:
27–29
Cruden, DM; Varnes, DJ (1996) ‘Landslide types and processes.’ In Turner, AK; Schuster, RL (eds), Landslides: Investigation and
Mitigation, Transportation Research Board Special Report 247, pp 37–75. Washington DC, USA: National Academies Press
Dekens, J (2007) Herders of Chitral: The lost messengers? Local knowledge on disaster preparedness in Chitral District, Pakistan.
Kathmandu, Nepal: ICIMOD
Dhital, MR; Khanal, N; Thapa, KB (1993) The role of extreme weather events, mass movements, and land use changes in
increasing natural hazards. Kathmandu, Nepal: ICIMOD
Dixit, A (2003) ‘Floods and vulnerability: Need to rethink flood management.’ Natural Hazards 28: 155–179
DSCWM (2004) Soil conservation and watershed management measures and low cost techniques. Kathmandu, Nepal: HMG
Nepal, Soil Conservation and Watershed Management Component (NARMSAP), Department of Soil Conservation and
Watershed Management; Danida
DSCWM (2005) Training handout on bioengineering and survey: Design and estimation of soil conservation activities.
Kathmandu, Nepal: HMG Nepal, Soil Conservation and Watershed Management Component (NARMSAP), Department of
Soil Conservation and Watershed Management; Danida
DWIDP/JICA (2004a) ‘Bioengineering guideline for hill slope stabilization.’ In Collection of guidelines regarding the project
period of DPTC/DMSP. Kathmandu, Nepal: Department of Water Induced Disaster Prevention (DWIDP), Disaster Mitigation
Support Programme Project (DMSP); JICA
83
DWIDP/JICA (2004b) ‘The basis principles of landslides, Training textbook.’ In Collection of guidelines regarding the project
period of DPTC/DMSP. Kathmandu, Nepal: Department of Water Induced Disaster Prevention (DWIDP), Disaster Mitigation
Support Programme Project (DMSP); JICA
EC (1997) Premo98: Prevention in the mountains for protection of the valleys: Principles and guidelines for the environmental
protection of drainage basins prone to flash floods. Experimental intervention in a pilot drainage basin (Section 3.3.3: River
training works). Brussels, Belgium: European Commission. http://ec.europa.eu/echo/civil_protection/civil/pdfdocs/pr98f02.
pdf (check dams and sills) http://ec.europa.eu/echo/civil_protection/civil/prote/cpactiv/cpact05g.htm (links to full report)
(accessed 21 June 2012)
Eriksson, M; Xu, J; Shrestha, AB; Vaidya, R; Nepal, S; Sandstrom, K (2009) The changing Himalayas : Impact of climate change
on water resources and livelihoods in the greater Himalayas. Kathmandu, Nepal: ICIMOD
FHWA (1989) Design of riprap revetment, Hydraulic Engineering Circular No. 11 (HEC-11 SI) Publication No. FHWAIP-89-016. McClean (VA), USA: Federal Highway Administration, Research, Development, and Technology. www.fhwa.dot.
gov/engineering/hydraulics/pubs/hec/hec11sI.pdf (converted document in metric units); http://isddc.dot.gov/OLPFiles/
FHWA/009881.pdf (original document with imperial and metric units) (accessed 27 June 2012)
FRM (2009) Flood Risk Management Act 2009: Scotland (asp 6). London and Edinburgh, UK: UK Government. www.legislation.
gov.uk/asp/2009/6/pdfs/asp_20090006_en.pdf (accessed 12 January 2012)
Geyik, MP (1986) Watershed management field manual – gully control, FAO Conservation Guide 13/2. Rome, Italy: Food and
Agriculture Organization
Grabs, WE; Hanisch, J (1993) ‘Objectives and prevention methods for glacier lake outburst floods (GLOFs).’ In Snow and glacier
hydrology: Proceedings of the Kathmandu symposium, November 1992, pp 341–352. Wallingford, UK: IAHS Press
Gray, DH; Sotir, RB (1996) Biotechnical and soil bioengineering slope stabilization: A practical guide for erosion control. New
York, USA: Wiley
GTZ and SDC (1999) Green roads in Nepal: Best practices report (2nd edn). Kathmandu, Nepal: German Technical Cooperation
and Swiss Agency for Development and Cooperation. www.scribd.com/doc/23231563/Green-Road-in-Nepal-Best-Practices
(accessed 23 April 2012)
Guedel, N (2008) ‘Gully plugging using check dams, WOCAT database QT NEP14.’ In Natural resource management
approaches and technologies in Nepal. Kathmandu, Nepal: ICIMOD
GWP (2012) What is IWRM? Stockholm, Sweden: Global Water Partnership. http://www.gwp.org/The-Challenge/What-is-IWRM/
(accessed 26 June 2012)
Hashemi Najafi, SFA; Ayoubzadeh, SA; Dehghani, AA (2008) ‘Experimental investigation of scour depth around L-head groynes
under clear water condition.’ Journal of Agricultural Sciences and Natural Resources 15(1): 192–203
Hellevang, K (2011) Sandbagging for flood protection. Fargo (ND), USA: North Dakota State University Extension Service. www.
ag.ndsu.edu/pubs/ageng/safety/ae626.pdf (accessed 19 June 2012)
HMGN (1990) Design manuals for irrigation projects in Nepal, M.7 Headworks, river training works and sedimentation manual.
Kathmandu, Nepal: His Majesty’s Government of Nepal, Ministry of Water Resources, Department of Irrigation, Planning and
Design Strengthening Project
HMGN (1999) Bio-engineering information, rate analysis norms (interim), standard specifications (interim), lists of species, and
supporting information. Kathmandu, Nepal: His Majesty’s Government of Nepal, Ministry of Works and Transportation,
Department of Road, Geo-environmental unit. www.dor.gov.np/documents/5.%20Bio-Engineering%20Information%20
(Rate%20Analysis%20Norms).pdf (accessed 11 July 2012)
Howell, J (1999) Roadside bio-engineering: Site handbook. Kathmandu, Nepal: His Majesty’s Government of Nepal,
Department of Roads
ICIMOD (1991a) Mountain risk engineering handbook, Volume 1. Kathmandu, Nepal: ICIMOD. www.icimod.org/publications/
index.php/search/publication/76 (accessed 15 July 2012)
ICIMOD (1991b) Mountain risk engineering handbook, Volume 2. Kathmandu, Nepal: ICIMOD. www.icimod.org/publications/
index.php/search/publication/77 (accessed 15 July 2012)
ICIMOD (2007) Good practices in watershed management. Kathmandu, Nepal: ICIMOD
ICIMOD (2008) Natural resource management approaches and technologies in Nepal, NEPCAT Fact Sheets. Kathmandu,
Nepal: ICIMOD
ICIMOD (2011) Glacial Lakes and glacial lake outburst floods in Nepal. Kathmandu, Nepal: ICIMOD
Ikeda, A (2004) ‘Design of sabo dam.’ In Collection of guidelines regarding to the project period of DPTC/DMSP. Kathmandu,
Nepal: Department of Water Induced Disaster Prevention (DWIDP); Disaster Mitigation Support Programme Project (DMSP);
JICA
84
References
Ives, JD; Shrestha, RB; Mool, PK (2010) Formation of glacial lakes in the Hindu Kush-Himalayas and GLOF risk assessment.
Kathmandu, Nepal: ICIMOD
Jha, H; Jha, S; Karmacharya, B (2000) Flood control measures, best practice report. An approach for community based flood
control measures in the Terai rivers. Kathmandu, Nepal: German Technical Cooperation (GTZ)
JICA (2006) ‘Technical guidance for landslide analysis and landslide mitigation measures in Nepal.’ In Collection of guidelines
regarding the project period of DPTC/DMSP. Kathmandu, Nepal: Department of Water Induced Disaster Prevention (DWIDP),
Disaster Mitigation Support Programme Project (DMSP); JICA
Julien, PY (2002) River mechanics. Cambridge, UK: Cambridge University Press
Kattelmann, R; Watanabe, T (1996) ‘Draining Himalayan glacial lakes before they burst.’ In Destructive water: Water-caused
natural disasters, their abatement and control, Proceedings of the Conference held at Anaheim, California, June 1996.
Wallingford, UK: IAHS Press
Klijn, F; Bruijn, K; Olfert, A; Penning-Rowsell, E; Simm, J; Wallis, M (2009) Flood risk assessment and flood risk management:
An introduction and guidance based on experiences and findings of FLOOD site (an EU-funded Integrated Project). Delft,
Netherlands: FLOODsite
Lammeranner, W; Rauch, HP; Laaha, G (2005) ‘Implementation and monitoring of soil bioengineering measures at a landslide
in the middle mountains of Nepal.’ Plant and Soil 278: 159–170
Li, G; Clarke, D (2007) ‘Vegetation patterns and the prediction of debris flow.’ In Chen, C; Major, JJ (eds), Debris-flow hazards
mitigation: Mechanics, prediction, and assessment, pp 35–44. Rotterdam, Netherlands: Mill Press
Maccaferri Australia (n.d.) Hydraulic structures: Groynes. Internet document, available at: www.maccaferri.com.au/
wawcs0137320/idDetails=213/Solutions-Groynes (accessed 20 April 2011)
Mal, BC (1999) Introduction to soil and water conservation engineering (surveying, irrigation, drainage and soil conservation).
New Delhi, India: Kalyani Publishers
McCullah, J; Gray, D (2005) Environmentally sensitive channel – and bank – protection measures, NCHRP Report 544.
Washington DC, USA: National Cooperative Highway Research Program, Transportation Research Board
MDEP (1997) Erosion and sediment control guidelines for urban and suburban areas (Riprap), pp 125–129. Boston (MA), USA:
Massachusetts Department of Environmental Protection. http://www.mass.gov/dep/water/essec2.pdf
MDEP (2003) Maine erosion and sediment control best management practices (BMPs) – E6 Riprap waterways. Augusta (ME),
USA: Maine Department of Environmental Protection, Bureau of Land and Water Quality. http://www.maine.gov/dep/land/
erosion/escbmps/escsectione6.pdf
Mehta, M (2007) ‘Gender and natural disasters: New challenges for mountain development.’ Disaster Preparedness 53: 17–19
Mool, PK (1995) ‘Glacier lake outburst floods in Nepal.’ Journal of Nepal Geological Society 11: 273–280
Mool, P; Bajracharya, SR; Joshi, SP (2001) Inventory of glaciers, glacier lakes and glacial lake outburst floods, monitoring and
early warning system in the Hindu Kush-Himalayan region. Kathmandu, Nepal: ICIMOD
NABARD (2012) Technical digest, Issue 14. Mumbai, India: National Bank for Agriculture and Rural Development, Technical
Services Department www.nabard.org/fileupload/DataBank/TechnicalDigest /Tech-Digest%20Issue%2014%20Jan%202012.
pdf (accessed 19 June 2012)
Nettleton, IM; Martin, S; Hencher, S; Moore, R (2005) ‘Debris flow types and mechanisms.’ In Winter, MG; Macgregor, F;
Shackman, L (eds), Scottish Road Network Landslides Study, pp 45–67. Edinburgh, UK: Scottish Government, the Scottish
Executive
ODNR (n.d.a) Live cribwalls, Ohio Stream Management Guide No. 17. Columbus, Ohio, USA: Ohio Department of Natural
Resources. www.dnr.state.oh.us/Portals/7/pubs/fs_st/stfs17.pdf (accessed 25 June 2012)
ODNR (n.d.b) Live fascines, Ohio Stream Management Guide No 14. Columbus, Ohio, USA: Ohio Department of Natural
Resources. www.dnr.state.oh.us/portals/7/pubs/fs_st/stfs14.pdf (accessed 25 June 2012)
PBC (2000) Riprap design and construction guide. Victoria (BC), Canada: Province of British Columbia, Ministry of Environment,
Lands and Parks www.env.gov.bc.ca/wsd/public_safety/flood/pdfs_word/riprap_guide.pdf (accessed 25 June 2012)
Rana, B; Shrestha, AB; Reynolds, JM; Aryal, R; Pokhrel, AP; Budhathoki, KP (2000) ‘Hazard assessment of the Tsho Rolpa glacier
lake and ongoing remediation measures.’ Journal of Nepal Geological Society 22: 563–570
Richardson SD; Reynolds JM (2000) ‘An overview of glacial hazards in the Himalayas.’ Quaternary International 65/66: 31–47
Safarzadeh, A; Salehi Neyshabouri, SAA; Ghodsian, M; Zarrati, AR (2010) ‘Experimental study of head shape effects on
shear stress distribution around a single groyne.’ In Dittrich, A; Koll, Ka; Aberle, J; Geisenhainer, P (eds), River flow 2010,
Proceedings of 5th International Conference on Fluvial Hydraulics, September 8–10, 2010, Braunschweig, Germany, pp
651–657. Karlsruhe, Germany: Bundesanstalt für Wasserbau. http://vzb.baw.de/digitale_bib/riverflow2010/PDF/A5/A5_02.
pdf (accessed 26 June 2012)
85
Schiechtl, HM; Stern, R (1992) Handbuch für naturnahen Erdbau. Eine Anleitung für ingenieurbiologische Bauweisen. Vienna,
Austria: Österreichischer Agrarverlag (German)
SDC (2008) Asia brief: Nepal – Roads to prosperity partnership results. Berne, Switzerland: SDC East Asia Division
Shang, Y; Yang, Z; Li, L; Liu, D; Liao, Q; Wang, Y (2003) ‘A super-large landslide in Tibet in 2000: Background, occurrence,
disaster, and origin.’ Geomorphology 54: 225–243
Sharda, VN; Juyal, GP; Prakash, C; Joshi, BP (2007) Soil conservation and watershed management: Training manual, Volume II.
Dehradun, India: Soil and Water Conservation Engineering, Central Soil &Water Conservation Research and Training Institute
Shrestha, AB (2008) Resource manual on flash flood risk management, Module 2: Non-Structural Measures. Kathmandu, Nepal:
ICIMOD
Shrestha, AB; Chapagain, PS; Thapa, R (2011) Flash flood risk management: A training of trainers manual. Kathmandu, Nepal:
ICIMOD
Singh, AK (2010) ‘Bioengineering techniques of slope stabilization and landslide mitigation.’ Disaster Prevention and
Management 19: 384–397. http://www.emeraldinsight.com/journals.htm?articleid=1865068&show=abstract (accessed 26
June 2012)
Singh, G (1980) Irrigation engineering. Delhi, India: Standard Book House
Singh, U; Wadhwani, AM; Johri, BM (1983) Dictionary of economic plants in India. New Delhi, India: Indian Council of
Agricultural Research
Sinha, PC (1998) Geological and mass movement disasters, Encyclopaedia of Disaster Management Series. New Delhi, India:
Anmol Publication
Suresh, R (1997) Soil and water conservation. New Delhi, India: Engineering Standard Publishers Distributors
Tabacchi, E; Lambs, L; Guilloy, H; Planty-Tabacchi, AM; Muller, E; Decamps, H (2000) ‘Impacts of riparian vegetation on
hydrological processes.’ Hydrological Processes 14: 2959–2976
Takahashi, T (2007) Debris flow: Mechanics, prediction and countermeasures. London, UK: Taylor & Francis
Tang, Y (1999) Manual on contour hedgerow intercropping technology. Kathmandu, Nepal: ICIMOD
Tang, Y; Murray AB (2004) Impact of contour hedgerows: A case study. Kathmandu, Nepal: ICIMOD
Tognacca, C; Bezzola, GR; Minor, HE (2000) ‘Threshold criterion for debris‐flow initiation due to channel‐bed failure.’ In
Wieczorek, GF; Naeser, ND (eds), Debris-flow hazards mitigation: Mechanics, prediction and assessment, pp 89‐97,
Rotterdam, Netherlands: Balkema
Uhlir, CF (1998) ‘Landslide-dammed lakes: A case study of the Lamabagar and Chaurikharka landslide deposits, Dolakha and
Solukhumbu districts, eastern Nepal.’ Journal of Nepal Geological Society 18: 329–334
UN-Water (2008) Status report on IWRM and water efficiency plans for CSD16. New York (NY), USA: UN-Water. www.unwater.
org/downloads/UNW_Status_Report_IWRM.pdf (accessed 24 April 2012)
Upadhya, M (2009) ‘Traditional technologies for water harvesting.’ ICIMOD, Sustainable Mountain Development (56): 24–26
USACE (1991) Hydrographic surveying, Engineer Manual EM 1110-2-1003. Washington DC, USA: U.S. Army Corps of
Engineers
Vaidya, RA (2009) ‘The role of water storage in adaptation to climate change in the HKH region.’ ICIMOD, Sustainable
Mountain Development (56): 10–13
Varnes, DJ, (1978) ‘Slope movement types and processes.’ In Schuster, RL; Krizek, RJ (eds), Landslides: Analysis and control,
Transportation Research Board Special Report 176, pp 11–33. Washington DC, USA: United States National Research
Council
Varshney, RS; Gupta, SC; Gupta, RL (1983) Theory and design of irrigation structures, Vol. I, Channels and tubewells (5th edn).
Roorkee, India: Nem Chand & Bros
Vazirani, VN; Ratwani, MM (1980) Analysis of structure: Analysis, design & detail of structures, Vol. II, (8th edn). Delhi, India:
Khanna Publishers
Vuichard, D; Zimmermann, M (1987) ‘The 1985 catastrophic drainage of a moraine-dammed lake, Khumbu Himal, Nepal:
Cause and consequences.’ Mountain Research and Development 7: 91–110
WHO/UNEP (1991) Surface water drainage for low income communities. Geneva, Switzerland: World Health Organization and
United Nations Environment Programme
WMO (2006a) Environmental aspects of integrated flood management, APFM Technical Document No 3, Flood Management
Policy Series. Geneva, Switzerland: World Meteorological Organization
WMO (2006b) Integrated flood management: Some environmental issues, WMO Bulletin Geneva, Switzerland: World
Meteorological Organization
86
References
Wu, J; Li, T (2001) ‘Behaviour and characteristics of debris flows.’ In Li Tianchi; Chalise SR; Upreti BN (eds), Landslide hazard
mitigation in the Hindu Kush-Himalayas, pp 203–214. Kathmandu, Nepal: ICIMOD
Xu, J; Shrestha, A; Vaidya, R; Eriksson, M; Hewitt, K (2007) The melting Himalayas: Regional challenges and local impacts of
climate change on mountain ecosystems and livelihoods. Kathmandu, Nepal: ICIMOD
Zhou, L; Wan, X (2005) Flash floods events and mitigation management in China. Capacity building for flash floods
management and sustainable development in the Himalayas: Status report of flash floods events and mitigation management
in China, a report submitted to ICIMOD, Kathmandu, Nepal
Zhu, P; Li T (2000) ‘Flash flooding caused by landslide dam failure.’ ICIMOD Newsletter 38: 4–5
Zhu, PY; Wang, CH; Wang, YC (2003) ‘Formation of an exceptionally serious outburst flood from a landslide dam in Tibet.’
Newsletter of Japanese Landslide Society 14/15: 23–25
Zhuang, J (2010) The 12 May Wenchuan earthquake-induced landslide lakes: Distribution and preliminary risk evaluation. Paper
presented at the workshop on State of the Art in Flash Flood Risk Management in the Himalayas, ICIMOD, 15–16 February
2010, Kathmandu, Nepal
87
90
© ICIMOD 2012
GPO Box 3226, Kathmandu, Nepal
Tel +977-1-5003222 Fax +977-1-5003299
Email info@icimod.org Web www.icimod.org
ISBN 978 92 9115 265 0